]>
Commit | Line | Data |
---|---|---|
7c673cae FG |
1 | #include <boost/utility/string_ref.hpp> |
2 | ||
3 | #include "common/ceph_json.h" | |
4 | #include "common/RWLock.h" | |
5 | #include "common/RefCountedObj.h" | |
6 | #include "common/WorkQueue.h" | |
7 | #include "common/Throttle.h" | |
8 | #include "common/errno.h" | |
9 | ||
10 | #include "rgw_common.h" | |
11 | #include "rgw_rados.h" | |
12 | #include "rgw_sync.h" | |
13 | #include "rgw_data_sync.h" | |
14 | #include "rgw_rest_conn.h" | |
15 | #include "rgw_cr_rados.h" | |
16 | #include "rgw_cr_rest.h" | |
17 | #include "rgw_http_client.h" | |
18 | #include "rgw_bucket.h" | |
19 | #include "rgw_metadata.h" | |
7c673cae FG |
20 | #include "rgw_sync_module.h" |
21 | ||
22 | #include "cls/lock/cls_lock_client.h" | |
23 | ||
31f18b77 FG |
24 | #include "auth/Crypto.h" |
25 | ||
26 | #include <boost/asio/yield.hpp> | |
27 | ||
7c673cae FG |
28 | #define dout_subsys ceph_subsys_rgw |
29 | ||
30 | #undef dout_prefix | |
31 | #define dout_prefix (*_dout << "data sync: ") | |
32 | ||
33 | static string datalog_sync_status_oid_prefix = "datalog.sync-status"; | |
34 | static string datalog_sync_status_shard_prefix = "datalog.sync-status.shard"; | |
35 | static string datalog_sync_full_sync_index_prefix = "data.full-sync.index"; | |
36 | static string bucket_status_oid_prefix = "bucket.sync-status"; | |
37 | ||
38 | class RGWSyncDebugLogger { | |
39 | CephContext *cct; | |
40 | string prefix; | |
41 | ||
42 | bool ended; | |
43 | ||
44 | public: | |
45 | RGWSyncDebugLogger(CephContext *_cct, const string& source_zone, | |
46 | const string& sync_type, const string& sync_stage, | |
47 | const string& resource, bool log_start = true) { | |
48 | init(_cct, source_zone, sync_type, sync_stage, resource, log_start); | |
49 | } | |
50 | RGWSyncDebugLogger() : cct(NULL), ended(false) {} | |
51 | ~RGWSyncDebugLogger(); | |
52 | ||
53 | void init(CephContext *_cct, const string& source_zone, | |
54 | const string& sync_type, const string& sync_stage, | |
55 | const string& resource, bool log_start = true); | |
56 | void log(const string& state); | |
57 | void finish(int status); | |
58 | }; | |
59 | ||
60 | void RGWSyncDebugLogger::init(CephContext *_cct, const string& source_zone, | |
61 | const string& sync_type, const string& sync_section, | |
62 | const string& resource, bool log_start) | |
63 | { | |
64 | cct = _cct; | |
65 | ended = false; | |
66 | string zone_str = source_zone.substr(0, 8); | |
67 | prefix = "Sync:" + zone_str + ":" + sync_type + ":" + sync_section + ":" + resource; | |
68 | if (log_start) { | |
69 | log("start"); | |
70 | } | |
71 | } | |
72 | ||
73 | RGWSyncDebugLogger::~RGWSyncDebugLogger() | |
74 | { | |
75 | if (!ended) { | |
76 | log("finish"); | |
77 | } | |
78 | } | |
79 | ||
80 | void RGWSyncDebugLogger::log(const string& state) | |
81 | { | |
82 | ldout(cct, 5) << prefix << ":" << state << dendl; | |
83 | } | |
84 | ||
85 | void RGWSyncDebugLogger::finish(int status) | |
86 | { | |
87 | ended = true; | |
88 | ldout(cct, 5) << prefix << ":" << "finish r=" << status << dendl; | |
89 | } | |
90 | ||
91 | class RGWDataSyncDebugLogger : public RGWSyncDebugLogger { | |
92 | public: | |
93 | RGWDataSyncDebugLogger() {} | |
94 | RGWDataSyncDebugLogger(RGWDataSyncEnv *sync_env, const string& sync_section, | |
95 | const string& resource, bool log_start = true) { | |
96 | init(sync_env, sync_section, resource, log_start); | |
97 | } | |
98 | void init(RGWDataSyncEnv *sync_env, const string& sync_section, | |
99 | const string& resource, bool log_start = true) { | |
100 | RGWSyncDebugLogger::init(sync_env->cct, sync_env->source_zone, "data", sync_section, resource, log_start); | |
101 | } | |
102 | ||
103 | }; | |
104 | ||
105 | void rgw_datalog_info::decode_json(JSONObj *obj) { | |
106 | JSONDecoder::decode_json("num_objects", num_shards, obj); | |
107 | } | |
108 | ||
109 | void rgw_datalog_entry::decode_json(JSONObj *obj) { | |
110 | JSONDecoder::decode_json("key", key, obj); | |
111 | utime_t ut; | |
112 | JSONDecoder::decode_json("timestamp", ut, obj); | |
113 | timestamp = ut.to_real_time(); | |
114 | } | |
115 | ||
116 | void rgw_datalog_shard_data::decode_json(JSONObj *obj) { | |
117 | JSONDecoder::decode_json("marker", marker, obj); | |
118 | JSONDecoder::decode_json("truncated", truncated, obj); | |
119 | JSONDecoder::decode_json("entries", entries, obj); | |
120 | }; | |
121 | ||
122 | class RGWReadDataSyncStatusMarkersCR : public RGWShardCollectCR { | |
123 | static constexpr int MAX_CONCURRENT_SHARDS = 16; | |
124 | ||
125 | RGWDataSyncEnv *env; | |
126 | const int num_shards; | |
127 | int shard_id{0};; | |
128 | ||
129 | map<uint32_t, rgw_data_sync_marker>& markers; | |
130 | ||
131 | public: | |
132 | RGWReadDataSyncStatusMarkersCR(RGWDataSyncEnv *env, int num_shards, | |
133 | map<uint32_t, rgw_data_sync_marker>& markers) | |
134 | : RGWShardCollectCR(env->cct, MAX_CONCURRENT_SHARDS), | |
135 | env(env), num_shards(num_shards), markers(markers) | |
136 | {} | |
137 | bool spawn_next() override; | |
138 | }; | |
139 | ||
140 | bool RGWReadDataSyncStatusMarkersCR::spawn_next() | |
141 | { | |
142 | if (shard_id >= num_shards) { | |
143 | return false; | |
144 | } | |
145 | using CR = RGWSimpleRadosReadCR<rgw_data_sync_marker>; | |
146 | spawn(new CR(env->async_rados, env->store, | |
147 | rgw_raw_obj(env->store->get_zone_params().log_pool, RGWDataSyncStatusManager::shard_obj_name(env->source_zone, shard_id)), | |
148 | &markers[shard_id]), | |
149 | false); | |
150 | shard_id++; | |
151 | return true; | |
152 | } | |
153 | ||
154 | class RGWReadDataSyncStatusCoroutine : public RGWCoroutine { | |
155 | RGWDataSyncEnv *sync_env; | |
156 | rgw_data_sync_status *sync_status; | |
157 | ||
158 | public: | |
159 | RGWReadDataSyncStatusCoroutine(RGWDataSyncEnv *_sync_env, | |
160 | rgw_data_sync_status *_status) | |
161 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), sync_status(_status) | |
162 | {} | |
163 | int operate() override; | |
164 | }; | |
165 | ||
166 | int RGWReadDataSyncStatusCoroutine::operate() | |
167 | { | |
168 | reenter(this) { | |
169 | // read sync info | |
170 | using ReadInfoCR = RGWSimpleRadosReadCR<rgw_data_sync_info>; | |
171 | yield { | |
172 | bool empty_on_enoent = false; // fail on ENOENT | |
173 | call(new ReadInfoCR(sync_env->async_rados, sync_env->store, | |
174 | rgw_raw_obj(sync_env->store->get_zone_params().log_pool, RGWDataSyncStatusManager::sync_status_oid(sync_env->source_zone)), | |
175 | &sync_status->sync_info, empty_on_enoent)); | |
176 | } | |
177 | if (retcode < 0) { | |
178 | ldout(sync_env->cct, 4) << "failed to read sync status info with " | |
179 | << cpp_strerror(retcode) << dendl; | |
180 | return set_cr_error(retcode); | |
181 | } | |
182 | // read shard markers | |
183 | using ReadMarkersCR = RGWReadDataSyncStatusMarkersCR; | |
184 | yield call(new ReadMarkersCR(sync_env, sync_status->sync_info.num_shards, | |
185 | sync_status->sync_markers)); | |
186 | if (retcode < 0) { | |
187 | ldout(sync_env->cct, 4) << "failed to read sync status markers with " | |
188 | << cpp_strerror(retcode) << dendl; | |
189 | return set_cr_error(retcode); | |
190 | } | |
191 | return set_cr_done(); | |
192 | } | |
193 | return 0; | |
194 | } | |
195 | ||
196 | class RGWReadRemoteDataLogShardInfoCR : public RGWCoroutine { | |
197 | RGWDataSyncEnv *sync_env; | |
198 | ||
199 | RGWRESTReadResource *http_op; | |
200 | ||
201 | int shard_id; | |
202 | RGWDataChangesLogInfo *shard_info; | |
203 | ||
204 | public: | |
205 | RGWReadRemoteDataLogShardInfoCR(RGWDataSyncEnv *_sync_env, | |
206 | int _shard_id, RGWDataChangesLogInfo *_shard_info) : RGWCoroutine(_sync_env->cct), | |
207 | sync_env(_sync_env), | |
208 | http_op(NULL), | |
209 | shard_id(_shard_id), | |
210 | shard_info(_shard_info) { | |
211 | } | |
212 | ||
213 | ~RGWReadRemoteDataLogShardInfoCR() override { | |
214 | if (http_op) { | |
215 | http_op->put(); | |
216 | } | |
217 | } | |
218 | ||
219 | int operate() override { | |
220 | reenter(this) { | |
221 | yield { | |
222 | char buf[16]; | |
223 | snprintf(buf, sizeof(buf), "%d", shard_id); | |
224 | rgw_http_param_pair pairs[] = { { "type" , "data" }, | |
225 | { "id", buf }, | |
226 | { "info" , NULL }, | |
227 | { NULL, NULL } }; | |
228 | ||
229 | string p = "/admin/log/"; | |
230 | ||
231 | http_op = new RGWRESTReadResource(sync_env->conn, p, pairs, NULL, sync_env->http_manager); | |
232 | ||
233 | http_op->set_user_info((void *)stack); | |
234 | ||
235 | int ret = http_op->aio_read(); | |
236 | if (ret < 0) { | |
237 | ldout(sync_env->cct, 0) << "ERROR: failed to read from " << p << dendl; | |
238 | log_error() << "failed to send http operation: " << http_op->to_str() << " ret=" << ret << std::endl; | |
239 | return set_cr_error(ret); | |
240 | } | |
241 | ||
242 | return io_block(0); | |
243 | } | |
244 | yield { | |
245 | int ret = http_op->wait(shard_info); | |
246 | if (ret < 0) { | |
247 | return set_cr_error(ret); | |
248 | } | |
249 | return set_cr_done(); | |
250 | } | |
251 | } | |
252 | return 0; | |
253 | } | |
254 | }; | |
255 | ||
256 | struct read_remote_data_log_response { | |
257 | string marker; | |
258 | bool truncated; | |
259 | list<rgw_data_change_log_entry> entries; | |
260 | ||
261 | read_remote_data_log_response() : truncated(false) {} | |
262 | ||
263 | void decode_json(JSONObj *obj) { | |
264 | JSONDecoder::decode_json("marker", marker, obj); | |
265 | JSONDecoder::decode_json("truncated", truncated, obj); | |
266 | JSONDecoder::decode_json("entries", entries, obj); | |
267 | }; | |
268 | }; | |
269 | ||
270 | class RGWReadRemoteDataLogShardCR : public RGWCoroutine { | |
271 | RGWDataSyncEnv *sync_env; | |
272 | ||
273 | RGWRESTReadResource *http_op; | |
274 | ||
275 | int shard_id; | |
276 | string *pmarker; | |
277 | list<rgw_data_change_log_entry> *entries; | |
278 | bool *truncated; | |
279 | ||
280 | read_remote_data_log_response response; | |
281 | ||
282 | public: | |
283 | RGWReadRemoteDataLogShardCR(RGWDataSyncEnv *_sync_env, | |
284 | int _shard_id, string *_pmarker, list<rgw_data_change_log_entry> *_entries, bool *_truncated) : RGWCoroutine(_sync_env->cct), | |
285 | sync_env(_sync_env), | |
286 | http_op(NULL), | |
287 | shard_id(_shard_id), | |
288 | pmarker(_pmarker), | |
289 | entries(_entries), | |
290 | truncated(_truncated) { | |
291 | } | |
292 | ~RGWReadRemoteDataLogShardCR() override { | |
293 | if (http_op) { | |
294 | http_op->put(); | |
295 | } | |
296 | } | |
297 | ||
298 | int operate() override { | |
299 | reenter(this) { | |
300 | yield { | |
301 | char buf[16]; | |
302 | snprintf(buf, sizeof(buf), "%d", shard_id); | |
303 | rgw_http_param_pair pairs[] = { { "type" , "data" }, | |
304 | { "id", buf }, | |
305 | { "marker", pmarker->c_str() }, | |
306 | { "extra-info", "true" }, | |
307 | { NULL, NULL } }; | |
308 | ||
309 | string p = "/admin/log/"; | |
310 | ||
311 | http_op = new RGWRESTReadResource(sync_env->conn, p, pairs, NULL, sync_env->http_manager); | |
312 | ||
313 | http_op->set_user_info((void *)stack); | |
314 | ||
315 | int ret = http_op->aio_read(); | |
316 | if (ret < 0) { | |
317 | ldout(sync_env->cct, 0) << "ERROR: failed to read from " << p << dendl; | |
318 | log_error() << "failed to send http operation: " << http_op->to_str() << " ret=" << ret << std::endl; | |
319 | return set_cr_error(ret); | |
320 | } | |
321 | ||
322 | return io_block(0); | |
323 | } | |
324 | yield { | |
325 | int ret = http_op->wait(&response); | |
326 | if (ret < 0) { | |
327 | return set_cr_error(ret); | |
328 | } | |
329 | entries->clear(); | |
330 | entries->swap(response.entries); | |
331 | *pmarker = response.marker; | |
332 | *truncated = response.truncated; | |
333 | return set_cr_done(); | |
334 | } | |
335 | } | |
336 | return 0; | |
337 | } | |
338 | }; | |
339 | ||
340 | class RGWReadRemoteDataLogInfoCR : public RGWShardCollectCR { | |
341 | RGWDataSyncEnv *sync_env; | |
342 | ||
343 | int num_shards; | |
344 | map<int, RGWDataChangesLogInfo> *datalog_info; | |
345 | ||
346 | int shard_id; | |
347 | #define READ_DATALOG_MAX_CONCURRENT 10 | |
348 | ||
349 | public: | |
350 | RGWReadRemoteDataLogInfoCR(RGWDataSyncEnv *_sync_env, | |
351 | int _num_shards, | |
352 | map<int, RGWDataChangesLogInfo> *_datalog_info) : RGWShardCollectCR(_sync_env->cct, READ_DATALOG_MAX_CONCURRENT), | |
353 | sync_env(_sync_env), num_shards(_num_shards), | |
354 | datalog_info(_datalog_info), shard_id(0) {} | |
355 | bool spawn_next() override; | |
356 | }; | |
357 | ||
358 | bool RGWReadRemoteDataLogInfoCR::spawn_next() { | |
359 | if (shard_id >= num_shards) { | |
360 | return false; | |
361 | } | |
362 | spawn(new RGWReadRemoteDataLogShardInfoCR(sync_env, shard_id, &(*datalog_info)[shard_id]), false); | |
363 | shard_id++; | |
364 | return true; | |
365 | } | |
366 | ||
367 | class RGWListRemoteDataLogShardCR : public RGWSimpleCoroutine { | |
368 | RGWDataSyncEnv *sync_env; | |
369 | RGWRESTReadResource *http_op; | |
370 | ||
371 | int shard_id; | |
372 | string marker; | |
373 | uint32_t max_entries; | |
374 | rgw_datalog_shard_data *result; | |
375 | ||
376 | public: | |
377 | RGWListRemoteDataLogShardCR(RGWDataSyncEnv *env, int _shard_id, | |
378 | const string& _marker, uint32_t _max_entries, | |
379 | rgw_datalog_shard_data *_result) | |
380 | : RGWSimpleCoroutine(env->store->ctx()), sync_env(env), http_op(NULL), | |
381 | shard_id(_shard_id), marker(_marker), max_entries(_max_entries), result(_result) {} | |
382 | ||
383 | int send_request() override { | |
384 | RGWRESTConn *conn = sync_env->conn; | |
385 | RGWRados *store = sync_env->store; | |
386 | ||
387 | char buf[32]; | |
388 | snprintf(buf, sizeof(buf), "%d", shard_id); | |
389 | ||
390 | char max_entries_buf[32]; | |
391 | snprintf(max_entries_buf, sizeof(max_entries_buf), "%d", (int)max_entries); | |
392 | ||
393 | const char *marker_key = (marker.empty() ? "" : "marker"); | |
394 | ||
395 | rgw_http_param_pair pairs[] = { { "type", "data" }, | |
396 | { "id", buf }, | |
397 | { "max-entries", max_entries_buf }, | |
398 | { marker_key, marker.c_str() }, | |
399 | { NULL, NULL } }; | |
400 | ||
401 | string p = "/admin/log/"; | |
402 | ||
403 | http_op = new RGWRESTReadResource(conn, p, pairs, NULL, sync_env->http_manager); | |
404 | http_op->set_user_info((void *)stack); | |
405 | ||
406 | int ret = http_op->aio_read(); | |
407 | if (ret < 0) { | |
408 | ldout(store->ctx(), 0) << "ERROR: failed to read from " << p << dendl; | |
409 | log_error() << "failed to send http operation: " << http_op->to_str() << " ret=" << ret << std::endl; | |
410 | http_op->put(); | |
411 | return ret; | |
412 | } | |
413 | ||
414 | return 0; | |
415 | } | |
416 | ||
417 | int request_complete() override { | |
418 | int ret = http_op->wait(result); | |
419 | http_op->put(); | |
420 | if (ret < 0 && ret != -ENOENT) { | |
421 | ldout(sync_env->store->ctx(), 0) << "ERROR: failed to list remote datalog shard, ret=" << ret << dendl; | |
422 | return ret; | |
423 | } | |
424 | return 0; | |
425 | } | |
426 | }; | |
427 | ||
428 | class RGWListRemoteDataLogCR : public RGWShardCollectCR { | |
429 | RGWDataSyncEnv *sync_env; | |
430 | ||
431 | map<int, string> shards; | |
432 | int max_entries_per_shard; | |
433 | map<int, rgw_datalog_shard_data> *result; | |
434 | ||
435 | map<int, string>::iterator iter; | |
436 | #define READ_DATALOG_MAX_CONCURRENT 10 | |
437 | ||
438 | public: | |
439 | RGWListRemoteDataLogCR(RGWDataSyncEnv *_sync_env, | |
440 | map<int, string>& _shards, | |
441 | int _max_entries_per_shard, | |
442 | map<int, rgw_datalog_shard_data> *_result) : RGWShardCollectCR(_sync_env->cct, READ_DATALOG_MAX_CONCURRENT), | |
443 | sync_env(_sync_env), max_entries_per_shard(_max_entries_per_shard), | |
444 | result(_result) { | |
445 | shards.swap(_shards); | |
446 | iter = shards.begin(); | |
447 | } | |
448 | bool spawn_next() override; | |
449 | }; | |
450 | ||
451 | bool RGWListRemoteDataLogCR::spawn_next() { | |
452 | if (iter == shards.end()) { | |
453 | return false; | |
454 | } | |
455 | ||
456 | spawn(new RGWListRemoteDataLogShardCR(sync_env, iter->first, iter->second, max_entries_per_shard, &(*result)[iter->first]), false); | |
457 | ++iter; | |
458 | return true; | |
459 | } | |
460 | ||
461 | class RGWInitDataSyncStatusCoroutine : public RGWCoroutine { | |
462 | static constexpr uint32_t lock_duration = 30; | |
463 | RGWDataSyncEnv *sync_env; | |
464 | RGWRados *store; | |
465 | const rgw_pool& pool; | |
466 | const uint32_t num_shards; | |
467 | ||
468 | string sync_status_oid; | |
469 | ||
470 | string lock_name; | |
471 | string cookie; | |
472 | rgw_data_sync_status *status; | |
473 | map<int, RGWDataChangesLogInfo> shards_info; | |
474 | public: | |
475 | RGWInitDataSyncStatusCoroutine(RGWDataSyncEnv *_sync_env, uint32_t num_shards, | |
31f18b77 | 476 | uint64_t instance_id, |
7c673cae FG |
477 | rgw_data_sync_status *status) |
478 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), store(sync_env->store), | |
479 | pool(store->get_zone_params().log_pool), | |
480 | num_shards(num_shards), status(status) { | |
481 | lock_name = "sync_lock"; | |
482 | ||
31f18b77 FG |
483 | status->sync_info.instance_id = instance_id; |
484 | ||
7c673cae FG |
485 | #define COOKIE_LEN 16 |
486 | char buf[COOKIE_LEN + 1]; | |
487 | ||
488 | gen_rand_alphanumeric(cct, buf, sizeof(buf) - 1); | |
489 | cookie = buf; | |
490 | ||
491 | sync_status_oid = RGWDataSyncStatusManager::sync_status_oid(sync_env->source_zone); | |
492 | } | |
493 | ||
494 | int operate() override { | |
495 | int ret; | |
496 | reenter(this) { | |
497 | using LockCR = RGWSimpleRadosLockCR; | |
498 | yield call(new LockCR(sync_env->async_rados, store, | |
499 | rgw_raw_obj{pool, sync_status_oid}, | |
500 | lock_name, cookie, lock_duration)); | |
501 | if (retcode < 0) { | |
502 | ldout(cct, 0) << "ERROR: failed to take a lock on " << sync_status_oid << dendl; | |
503 | return set_cr_error(retcode); | |
504 | } | |
505 | using WriteInfoCR = RGWSimpleRadosWriteCR<rgw_data_sync_info>; | |
506 | yield call(new WriteInfoCR(sync_env->async_rados, store, | |
507 | rgw_raw_obj{pool, sync_status_oid}, | |
508 | status->sync_info)); | |
509 | if (retcode < 0) { | |
510 | ldout(cct, 0) << "ERROR: failed to write sync status info with " << retcode << dendl; | |
511 | return set_cr_error(retcode); | |
512 | } | |
513 | ||
514 | /* take lock again, we just recreated the object */ | |
515 | yield call(new LockCR(sync_env->async_rados, store, | |
516 | rgw_raw_obj{pool, sync_status_oid}, | |
517 | lock_name, cookie, lock_duration)); | |
518 | if (retcode < 0) { | |
519 | ldout(cct, 0) << "ERROR: failed to take a lock on " << sync_status_oid << dendl; | |
520 | return set_cr_error(retcode); | |
521 | } | |
522 | ||
523 | /* fetch current position in logs */ | |
524 | yield { | |
525 | RGWRESTConn *conn = store->get_zone_conn_by_id(sync_env->source_zone); | |
526 | if (!conn) { | |
527 | ldout(cct, 0) << "ERROR: connection to zone " << sync_env->source_zone << " does not exist!" << dendl; | |
528 | return set_cr_error(-EIO); | |
529 | } | |
530 | for (uint32_t i = 0; i < num_shards; i++) { | |
531 | spawn(new RGWReadRemoteDataLogShardInfoCR(sync_env, i, &shards_info[i]), true); | |
532 | } | |
533 | } | |
534 | while (collect(&ret, NULL)) { | |
535 | if (ret < 0) { | |
536 | ldout(cct, 0) << "ERROR: failed to read remote data log shards" << dendl; | |
537 | return set_state(RGWCoroutine_Error); | |
538 | } | |
539 | yield; | |
540 | } | |
541 | yield { | |
542 | for (uint32_t i = 0; i < num_shards; i++) { | |
543 | RGWDataChangesLogInfo& info = shards_info[i]; | |
544 | auto& marker = status->sync_markers[i]; | |
545 | marker.next_step_marker = info.marker; | |
546 | marker.timestamp = info.last_update; | |
547 | const auto& oid = RGWDataSyncStatusManager::shard_obj_name(sync_env->source_zone, i); | |
548 | using WriteMarkerCR = RGWSimpleRadosWriteCR<rgw_data_sync_marker>; | |
549 | spawn(new WriteMarkerCR(sync_env->async_rados, store, | |
550 | rgw_raw_obj{pool, oid}, marker), true); | |
551 | } | |
552 | } | |
553 | while (collect(&ret, NULL)) { | |
554 | if (ret < 0) { | |
555 | ldout(cct, 0) << "ERROR: failed to write data sync status markers" << dendl; | |
556 | return set_state(RGWCoroutine_Error); | |
557 | } | |
558 | yield; | |
559 | } | |
560 | ||
561 | status->sync_info.state = rgw_data_sync_info::StateBuildingFullSyncMaps; | |
562 | yield call(new WriteInfoCR(sync_env->async_rados, store, | |
563 | rgw_raw_obj{pool, sync_status_oid}, | |
564 | status->sync_info)); | |
565 | if (retcode < 0) { | |
566 | ldout(cct, 0) << "ERROR: failed to write sync status info with " << retcode << dendl; | |
567 | return set_cr_error(retcode); | |
568 | } | |
569 | yield call(new RGWSimpleRadosUnlockCR(sync_env->async_rados, store, | |
570 | rgw_raw_obj{pool, sync_status_oid}, | |
571 | lock_name, cookie)); | |
572 | return set_cr_done(); | |
573 | } | |
574 | return 0; | |
575 | } | |
576 | }; | |
577 | ||
578 | int RGWRemoteDataLog::read_log_info(rgw_datalog_info *log_info) | |
579 | { | |
580 | rgw_http_param_pair pairs[] = { { "type", "data" }, | |
581 | { NULL, NULL } }; | |
582 | ||
583 | int ret = sync_env.conn->get_json_resource("/admin/log", pairs, *log_info); | |
584 | if (ret < 0) { | |
585 | ldout(store->ctx(), 0) << "ERROR: failed to fetch datalog info" << dendl; | |
586 | return ret; | |
587 | } | |
588 | ||
589 | ldout(store->ctx(), 20) << "remote datalog, num_shards=" << log_info->num_shards << dendl; | |
590 | ||
591 | return 0; | |
592 | } | |
593 | ||
594 | int RGWRemoteDataLog::read_source_log_shards_info(map<int, RGWDataChangesLogInfo> *shards_info) | |
595 | { | |
596 | rgw_datalog_info log_info; | |
597 | int ret = read_log_info(&log_info); | |
598 | if (ret < 0) { | |
599 | return ret; | |
600 | } | |
601 | ||
602 | return run(new RGWReadRemoteDataLogInfoCR(&sync_env, log_info.num_shards, shards_info)); | |
603 | } | |
604 | ||
605 | int RGWRemoteDataLog::read_source_log_shards_next(map<int, string> shard_markers, map<int, rgw_datalog_shard_data> *result) | |
606 | { | |
607 | if (store->is_meta_master()) { | |
608 | return 0; | |
609 | } | |
610 | ||
611 | return run(new RGWListRemoteDataLogCR(&sync_env, shard_markers, 1, result)); | |
612 | } | |
613 | ||
614 | int RGWRemoteDataLog::init(const string& _source_zone, RGWRESTConn *_conn, RGWSyncErrorLogger *_error_logger, RGWSyncModuleInstanceRef& _sync_module) | |
615 | { | |
616 | sync_env.init(store->ctx(), store, _conn, async_rados, &http_manager, _error_logger, _source_zone, _sync_module); | |
617 | ||
618 | if (initialized) { | |
619 | return 0; | |
620 | } | |
621 | ||
622 | int ret = http_manager.set_threaded(); | |
623 | if (ret < 0) { | |
624 | ldout(store->ctx(), 0) << "failed in http_manager.set_threaded() ret=" << ret << dendl; | |
625 | return ret; | |
626 | } | |
627 | ||
628 | initialized = true; | |
629 | ||
630 | return 0; | |
631 | } | |
632 | ||
633 | void RGWRemoteDataLog::finish() | |
634 | { | |
635 | stop(); | |
636 | } | |
637 | ||
638 | int RGWRemoteDataLog::get_shard_info(int shard_id) | |
639 | { | |
640 | char buf[32]; | |
641 | snprintf(buf, sizeof(buf), "%d", shard_id); | |
642 | ||
643 | rgw_http_param_pair pairs[] = { { "type", "data" }, | |
644 | { "id", buf }, | |
645 | { "info", NULL }, | |
646 | { NULL, NULL } }; | |
647 | ||
648 | RGWDataChangesLogInfo info; | |
649 | int ret = sync_env.conn->get_json_resource("/admin/log", pairs, info); | |
650 | if (ret < 0) { | |
651 | ldout(store->ctx(), 0) << "ERROR: failed to fetch datalog info" << dendl; | |
652 | return ret; | |
653 | } | |
654 | ||
655 | ldout(store->ctx(), 20) << "remote datalog, shard_id=" << shard_id << " marker=" << info.marker << dendl; | |
656 | ||
657 | return 0; | |
658 | } | |
659 | ||
660 | int RGWRemoteDataLog::read_sync_status(rgw_data_sync_status *sync_status) | |
661 | { | |
662 | // cannot run concurrently with run_sync(), so run in a separate manager | |
663 | RGWCoroutinesManager crs(store->ctx(), store->get_cr_registry()); | |
664 | RGWHTTPManager http_manager(store->ctx(), crs.get_completion_mgr()); | |
665 | int ret = http_manager.set_threaded(); | |
666 | if (ret < 0) { | |
667 | ldout(store->ctx(), 0) << "failed in http_manager.set_threaded() ret=" << ret << dendl; | |
668 | return ret; | |
669 | } | |
670 | RGWDataSyncEnv sync_env_local = sync_env; | |
671 | sync_env_local.http_manager = &http_manager; | |
672 | ret = crs.run(new RGWReadDataSyncStatusCoroutine(&sync_env_local, sync_status)); | |
673 | http_manager.stop(); | |
674 | return ret; | |
675 | } | |
676 | ||
677 | int RGWRemoteDataLog::init_sync_status(int num_shards) | |
678 | { | |
679 | rgw_data_sync_status sync_status; | |
680 | RGWCoroutinesManager crs(store->ctx(), store->get_cr_registry()); | |
681 | RGWHTTPManager http_manager(store->ctx(), crs.get_completion_mgr()); | |
682 | int ret = http_manager.set_threaded(); | |
683 | if (ret < 0) { | |
684 | ldout(store->ctx(), 0) << "failed in http_manager.set_threaded() ret=" << ret << dendl; | |
685 | return ret; | |
686 | } | |
687 | RGWDataSyncEnv sync_env_local = sync_env; | |
688 | sync_env_local.http_manager = &http_manager; | |
31f18b77 FG |
689 | uint64_t instance_id; |
690 | get_random_bytes((char *)&instance_id, sizeof(instance_id)); | |
691 | ret = crs.run(new RGWInitDataSyncStatusCoroutine(&sync_env_local, num_shards, instance_id, &sync_status)); | |
7c673cae FG |
692 | http_manager.stop(); |
693 | return ret; | |
694 | } | |
695 | ||
696 | static string full_data_sync_index_shard_oid(const string& source_zone, int shard_id) | |
697 | { | |
698 | char buf[datalog_sync_full_sync_index_prefix.size() + 1 + source_zone.size() + 1 + 16]; | |
699 | snprintf(buf, sizeof(buf), "%s.%s.%d", datalog_sync_full_sync_index_prefix.c_str(), source_zone.c_str(), shard_id); | |
700 | return string(buf); | |
701 | } | |
702 | ||
703 | struct bucket_instance_meta_info { | |
704 | string key; | |
705 | obj_version ver; | |
706 | utime_t mtime; | |
707 | RGWBucketInstanceMetadataObject data; | |
708 | ||
709 | bucket_instance_meta_info() {} | |
710 | ||
711 | void decode_json(JSONObj *obj) { | |
712 | JSONDecoder::decode_json("key", key, obj); | |
713 | JSONDecoder::decode_json("ver", ver, obj); | |
714 | JSONDecoder::decode_json("mtime", mtime, obj); | |
715 | JSONDecoder::decode_json("data", data, obj); | |
716 | } | |
717 | }; | |
718 | ||
719 | class RGWListBucketIndexesCR : public RGWCoroutine { | |
720 | RGWDataSyncEnv *sync_env; | |
721 | ||
722 | RGWRados *store; | |
723 | ||
724 | rgw_data_sync_status *sync_status; | |
725 | int num_shards; | |
726 | ||
727 | int req_ret; | |
728 | int ret; | |
729 | ||
730 | list<string> result; | |
731 | list<string>::iterator iter; | |
732 | ||
733 | RGWShardedOmapCRManager *entries_index; | |
734 | ||
735 | string oid_prefix; | |
736 | ||
737 | string path; | |
738 | bucket_instance_meta_info meta_info; | |
739 | string key; | |
740 | string s; | |
741 | int i; | |
742 | ||
743 | bool failed; | |
744 | ||
745 | public: | |
746 | RGWListBucketIndexesCR(RGWDataSyncEnv *_sync_env, | |
747 | rgw_data_sync_status *_sync_status) : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), | |
748 | store(sync_env->store), sync_status(_sync_status), | |
749 | req_ret(0), ret(0), entries_index(NULL), i(0), failed(false) { | |
750 | oid_prefix = datalog_sync_full_sync_index_prefix + "." + sync_env->source_zone; | |
751 | path = "/admin/metadata/bucket.instance"; | |
752 | num_shards = sync_status->sync_info.num_shards; | |
753 | } | |
754 | ~RGWListBucketIndexesCR() override { | |
755 | delete entries_index; | |
756 | } | |
757 | ||
758 | int operate() override { | |
759 | reenter(this) { | |
760 | entries_index = new RGWShardedOmapCRManager(sync_env->async_rados, store, this, num_shards, | |
761 | store->get_zone_params().log_pool, | |
762 | oid_prefix); | |
763 | yield { | |
764 | string entrypoint = string("/admin/metadata/bucket.instance"); | |
765 | /* FIXME: need a better scaling solution here, requires streaming output */ | |
766 | call(new RGWReadRESTResourceCR<list<string> >(store->ctx(), sync_env->conn, sync_env->http_manager, | |
767 | entrypoint, NULL, &result)); | |
768 | } | |
769 | if (get_ret_status() < 0) { | |
770 | ldout(sync_env->cct, 0) << "ERROR: failed to fetch metadata for section bucket.index" << dendl; | |
771 | return set_state(RGWCoroutine_Error); | |
772 | } | |
773 | for (iter = result.begin(); iter != result.end(); ++iter) { | |
774 | ldout(sync_env->cct, 20) << "list metadata: section=bucket.index key=" << *iter << dendl; | |
775 | ||
776 | key = *iter; | |
777 | ||
778 | yield { | |
779 | rgw_http_param_pair pairs[] = { { "key", key.c_str() }, | |
780 | { NULL, NULL } }; | |
781 | ||
782 | call(new RGWReadRESTResourceCR<bucket_instance_meta_info>(store->ctx(), sync_env->conn, sync_env->http_manager, path, pairs, &meta_info)); | |
783 | } | |
784 | ||
785 | num_shards = meta_info.data.get_bucket_info().num_shards; | |
786 | if (num_shards > 0) { | |
787 | for (i = 0; i < num_shards; i++) { | |
788 | char buf[16]; | |
789 | snprintf(buf, sizeof(buf), ":%d", i); | |
790 | s = key + buf; | |
791 | yield entries_index->append(s, store->data_log->get_log_shard_id(meta_info.data.get_bucket_info().bucket, i)); | |
792 | } | |
793 | } else { | |
794 | yield entries_index->append(key, store->data_log->get_log_shard_id(meta_info.data.get_bucket_info().bucket, -1)); | |
795 | } | |
796 | } | |
797 | yield { | |
798 | if (!entries_index->finish()) { | |
799 | failed = true; | |
800 | } | |
801 | } | |
802 | if (!failed) { | |
803 | for (map<uint32_t, rgw_data_sync_marker>::iterator iter = sync_status->sync_markers.begin(); iter != sync_status->sync_markers.end(); ++iter) { | |
804 | int shard_id = (int)iter->first; | |
805 | rgw_data_sync_marker& marker = iter->second; | |
806 | marker.total_entries = entries_index->get_total_entries(shard_id); | |
807 | spawn(new RGWSimpleRadosWriteCR<rgw_data_sync_marker>(sync_env->async_rados, store, | |
808 | rgw_raw_obj(store->get_zone_params().log_pool, RGWDataSyncStatusManager::shard_obj_name(sync_env->source_zone, shard_id)), | |
809 | marker), true); | |
810 | } | |
811 | } else { | |
812 | yield call(sync_env->error_logger->log_error_cr(sync_env->conn->get_remote_id(), "data.init", "", | |
813 | EIO, string("failed to build bucket instances map"))); | |
814 | } | |
815 | while (collect(&ret, NULL)) { | |
816 | if (ret < 0) { | |
817 | yield call(sync_env->error_logger->log_error_cr(sync_env->conn->get_remote_id(), "data.init", "", | |
818 | -ret, string("failed to store sync status: ") + cpp_strerror(-ret))); | |
819 | req_ret = ret; | |
820 | } | |
821 | yield; | |
822 | } | |
823 | drain_all(); | |
824 | if (req_ret < 0) { | |
825 | yield return set_cr_error(req_ret); | |
826 | } | |
827 | yield return set_cr_done(); | |
828 | } | |
829 | return 0; | |
830 | } | |
831 | }; | |
832 | ||
833 | #define DATA_SYNC_UPDATE_MARKER_WINDOW 1 | |
834 | ||
835 | class RGWDataSyncShardMarkerTrack : public RGWSyncShardMarkerTrack<string, string> { | |
836 | RGWDataSyncEnv *sync_env; | |
837 | ||
838 | string marker_oid; | |
839 | rgw_data_sync_marker sync_marker; | |
840 | ||
841 | map<string, string> key_to_marker; | |
842 | map<string, string> marker_to_key; | |
843 | ||
844 | void handle_finish(const string& marker) override { | |
845 | map<string, string>::iterator iter = marker_to_key.find(marker); | |
846 | if (iter == marker_to_key.end()) { | |
847 | return; | |
848 | } | |
849 | key_to_marker.erase(iter->second); | |
850 | reset_need_retry(iter->second); | |
851 | marker_to_key.erase(iter); | |
852 | } | |
853 | ||
854 | public: | |
855 | RGWDataSyncShardMarkerTrack(RGWDataSyncEnv *_sync_env, | |
856 | const string& _marker_oid, | |
857 | const rgw_data_sync_marker& _marker) : RGWSyncShardMarkerTrack(DATA_SYNC_UPDATE_MARKER_WINDOW), | |
858 | sync_env(_sync_env), | |
859 | marker_oid(_marker_oid), | |
860 | sync_marker(_marker) {} | |
861 | ||
862 | RGWCoroutine *store_marker(const string& new_marker, uint64_t index_pos, const real_time& timestamp) override { | |
863 | sync_marker.marker = new_marker; | |
864 | sync_marker.pos = index_pos; | |
865 | ||
866 | ldout(sync_env->cct, 20) << __func__ << "(): updating marker marker_oid=" << marker_oid << " marker=" << new_marker << dendl; | |
867 | RGWRados *store = sync_env->store; | |
868 | ||
869 | return new RGWSimpleRadosWriteCR<rgw_data_sync_marker>(sync_env->async_rados, store, | |
870 | rgw_raw_obj(store->get_zone_params().log_pool, marker_oid), | |
871 | sync_marker); | |
872 | } | |
873 | ||
874 | /* | |
875 | * create index from key -> marker, and from marker -> key | |
876 | * this is useful so that we can insure that we only have one | |
877 | * entry for any key that is used. This is needed when doing | |
878 | * incremenatl sync of data, and we don't want to run multiple | |
879 | * concurrent sync operations for the same bucket shard | |
880 | */ | |
881 | bool index_key_to_marker(const string& key, const string& marker) { | |
882 | if (key_to_marker.find(key) != key_to_marker.end()) { | |
883 | set_need_retry(key); | |
884 | return false; | |
885 | } | |
886 | key_to_marker[key] = marker; | |
887 | marker_to_key[marker] = key; | |
888 | return true; | |
889 | } | |
890 | }; | |
891 | ||
892 | // ostream wrappers to print buckets without copying strings | |
893 | struct bucket_str { | |
894 | const rgw_bucket& b; | |
895 | bucket_str(const rgw_bucket& b) : b(b) {} | |
896 | }; | |
897 | std::ostream& operator<<(std::ostream& out, const bucket_str& rhs) { | |
898 | auto& b = rhs.b; | |
899 | if (!b.tenant.empty()) { | |
900 | out << b.tenant << '/'; | |
901 | } | |
902 | out << b.name; | |
903 | if (!b.bucket_id.empty()) { | |
904 | out << ':' << b.bucket_id; | |
905 | } | |
906 | return out; | |
907 | } | |
908 | ||
909 | struct bucket_shard_str { | |
910 | const rgw_bucket_shard& bs; | |
911 | bucket_shard_str(const rgw_bucket_shard& bs) : bs(bs) {} | |
912 | }; | |
913 | std::ostream& operator<<(std::ostream& out, const bucket_shard_str& rhs) { | |
914 | auto& bs = rhs.bs; | |
915 | out << bucket_str{bs.bucket}; | |
916 | if (bs.shard_id >= 0) { | |
917 | out << ':' << bs.shard_id; | |
918 | } | |
919 | return out; | |
920 | } | |
921 | ||
922 | class RGWRunBucketSyncCoroutine : public RGWCoroutine { | |
923 | RGWDataSyncEnv *sync_env; | |
924 | rgw_bucket_shard bs; | |
925 | RGWBucketInfo bucket_info; | |
926 | rgw_bucket_shard_sync_info sync_status; | |
927 | RGWMetaSyncEnv meta_sync_env; | |
928 | ||
929 | RGWDataSyncDebugLogger logger; | |
930 | const std::string status_oid; | |
931 | ||
932 | boost::intrusive_ptr<RGWContinuousLeaseCR> lease_cr; | |
933 | boost::intrusive_ptr<RGWCoroutinesStack> lease_stack; | |
934 | ||
935 | public: | |
936 | RGWRunBucketSyncCoroutine(RGWDataSyncEnv *_sync_env, const rgw_bucket_shard& bs) | |
937 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), bs(bs), | |
938 | status_oid(RGWBucketSyncStatusManager::status_oid(sync_env->source_zone, bs)) { | |
939 | logger.init(sync_env, "Bucket", bs.get_key()); | |
940 | } | |
941 | ~RGWRunBucketSyncCoroutine() override { | |
942 | if (lease_cr) { | |
943 | lease_cr->abort(); | |
944 | } | |
945 | } | |
946 | ||
947 | int operate() override; | |
948 | }; | |
949 | ||
950 | class RGWDataSyncSingleEntryCR : public RGWCoroutine { | |
951 | RGWDataSyncEnv *sync_env; | |
952 | ||
953 | string raw_key; | |
954 | string entry_marker; | |
955 | ||
956 | rgw_bucket_shard bs; | |
957 | ||
958 | int sync_status; | |
959 | ||
960 | bufferlist md_bl; | |
961 | ||
962 | RGWDataSyncShardMarkerTrack *marker_tracker; | |
963 | ||
964 | boost::intrusive_ptr<RGWOmapAppend> error_repo; | |
965 | bool remove_from_repo; | |
966 | ||
967 | set<string> keys; | |
968 | ||
969 | public: | |
970 | RGWDataSyncSingleEntryCR(RGWDataSyncEnv *_sync_env, | |
971 | const string& _raw_key, const string& _entry_marker, RGWDataSyncShardMarkerTrack *_marker_tracker, | |
972 | RGWOmapAppend *_error_repo, bool _remove_from_repo) : RGWCoroutine(_sync_env->cct), | |
973 | sync_env(_sync_env), | |
974 | raw_key(_raw_key), entry_marker(_entry_marker), | |
975 | sync_status(0), | |
976 | marker_tracker(_marker_tracker), | |
977 | error_repo(_error_repo), remove_from_repo(_remove_from_repo) { | |
978 | set_description() << "data sync single entry (source_zone=" << sync_env->source_zone << ") key=" <<_raw_key << " entry=" << entry_marker; | |
979 | } | |
980 | ||
981 | int operate() override { | |
982 | reenter(this) { | |
983 | do { | |
984 | yield { | |
985 | int ret = rgw_bucket_parse_bucket_key(sync_env->cct, raw_key, | |
986 | &bs.bucket, &bs.shard_id); | |
987 | if (ret < 0) { | |
988 | return set_cr_error(-EIO); | |
989 | } | |
990 | if (marker_tracker) { | |
991 | marker_tracker->reset_need_retry(raw_key); | |
992 | } | |
993 | call(new RGWRunBucketSyncCoroutine(sync_env, bs)); | |
994 | } | |
995 | } while (marker_tracker && marker_tracker->need_retry(raw_key)); | |
996 | ||
997 | sync_status = retcode; | |
998 | ||
999 | if (sync_status == -ENOENT) { | |
1000 | // this was added when 'tenant/' was added to datalog entries, because | |
1001 | // preexisting tenant buckets could never sync and would stay in the | |
1002 | // error_repo forever | |
1003 | ldout(sync_env->store->ctx(), 0) << "WARNING: skipping data log entry " | |
1004 | "for missing bucket " << raw_key << dendl; | |
1005 | sync_status = 0; | |
1006 | } | |
1007 | ||
1008 | if (sync_status < 0) { | |
1009 | yield call(sync_env->error_logger->log_error_cr(sync_env->conn->get_remote_id(), "data", raw_key, | |
1010 | -sync_status, string("failed to sync bucket instance: ") + cpp_strerror(-sync_status))); | |
1011 | if (retcode < 0) { | |
1012 | ldout(sync_env->store->ctx(), 0) << "ERROR: failed to log sync failure: retcode=" << retcode << dendl; | |
1013 | } | |
1014 | if (error_repo && !error_repo->append(raw_key)) { | |
1015 | ldout(sync_env->store->ctx(), 0) << "ERROR: failed to log sync failure in error repo: retcode=" << retcode << dendl; | |
1016 | } | |
1017 | } else if (error_repo && remove_from_repo) { | |
1018 | keys = {raw_key}; | |
1019 | yield call(new RGWRadosRemoveOmapKeysCR(sync_env->store, error_repo->get_obj(), keys)); | |
1020 | if (retcode < 0) { | |
1021 | ldout(sync_env->store->ctx(), 0) << "ERROR: failed to remove omap key from error repo (" | |
1022 | << error_repo->get_obj() << " retcode=" << retcode << dendl; | |
1023 | } | |
1024 | } | |
1025 | /* FIXME: what do do in case of error */ | |
1026 | if (marker_tracker && !entry_marker.empty()) { | |
1027 | /* update marker */ | |
1028 | yield call(marker_tracker->finish(entry_marker)); | |
1029 | } | |
1030 | if (sync_status == 0) { | |
1031 | sync_status = retcode; | |
1032 | } | |
1033 | if (sync_status < 0) { | |
1034 | return set_cr_error(sync_status); | |
1035 | } | |
1036 | return set_cr_done(); | |
1037 | } | |
1038 | return 0; | |
1039 | } | |
1040 | }; | |
1041 | ||
1042 | #define BUCKET_SHARD_SYNC_SPAWN_WINDOW 20 | |
1043 | #define DATA_SYNC_MAX_ERR_ENTRIES 10 | |
1044 | ||
1045 | class RGWDataSyncShardCR : public RGWCoroutine { | |
1046 | RGWDataSyncEnv *sync_env; | |
1047 | ||
1048 | rgw_pool pool; | |
1049 | ||
1050 | uint32_t shard_id; | |
1051 | rgw_data_sync_marker sync_marker; | |
1052 | ||
1053 | map<string, bufferlist> entries; | |
1054 | map<string, bufferlist>::iterator iter; | |
1055 | ||
1056 | string oid; | |
1057 | ||
1058 | RGWDataSyncShardMarkerTrack *marker_tracker; | |
1059 | ||
1060 | list<rgw_data_change_log_entry> log_entries; | |
1061 | list<rgw_data_change_log_entry>::iterator log_iter; | |
1062 | bool truncated; | |
1063 | ||
1064 | RGWDataChangesLogInfo shard_info; | |
1065 | string datalog_marker; | |
1066 | ||
1067 | Mutex inc_lock; | |
1068 | Cond inc_cond; | |
1069 | ||
1070 | boost::asio::coroutine incremental_cr; | |
1071 | boost::asio::coroutine full_cr; | |
1072 | ||
1073 | ||
1074 | set<string> modified_shards; | |
1075 | set<string> current_modified; | |
1076 | ||
1077 | set<string>::iterator modified_iter; | |
1078 | ||
1079 | int total_entries; | |
1080 | ||
1081 | int spawn_window; | |
1082 | ||
1083 | bool *reset_backoff; | |
1084 | ||
1085 | set<string> spawned_keys; | |
1086 | ||
31f18b77 FG |
1087 | boost::intrusive_ptr<RGWContinuousLeaseCR> lease_cr; |
1088 | boost::intrusive_ptr<RGWCoroutinesStack> lease_stack; | |
7c673cae FG |
1089 | string status_oid; |
1090 | ||
1091 | ||
1092 | string error_oid; | |
1093 | RGWOmapAppend *error_repo; | |
1094 | map<string, bufferlist> error_entries; | |
1095 | string error_marker; | |
1096 | int max_error_entries; | |
1097 | ||
1098 | ceph::real_time error_retry_time; | |
1099 | ||
1100 | #define RETRY_BACKOFF_SECS_MIN 60 | |
1101 | #define RETRY_BACKOFF_SECS_DEFAULT 60 | |
1102 | #define RETRY_BACKOFF_SECS_MAX 600 | |
1103 | uint32_t retry_backoff_secs; | |
1104 | ||
1105 | RGWDataSyncDebugLogger logger; | |
1106 | public: | |
1107 | RGWDataSyncShardCR(RGWDataSyncEnv *_sync_env, | |
1108 | rgw_pool& _pool, | |
1109 | uint32_t _shard_id, rgw_data_sync_marker& _marker, bool *_reset_backoff) : RGWCoroutine(_sync_env->cct), | |
1110 | sync_env(_sync_env), | |
1111 | pool(_pool), | |
1112 | shard_id(_shard_id), | |
1113 | sync_marker(_marker), | |
1114 | marker_tracker(NULL), truncated(false), inc_lock("RGWDataSyncShardCR::inc_lock"), | |
1115 | total_entries(0), spawn_window(BUCKET_SHARD_SYNC_SPAWN_WINDOW), reset_backoff(NULL), | |
1116 | lease_cr(nullptr), lease_stack(nullptr), error_repo(nullptr), max_error_entries(DATA_SYNC_MAX_ERR_ENTRIES), | |
1117 | retry_backoff_secs(RETRY_BACKOFF_SECS_DEFAULT) { | |
1118 | set_description() << "data sync shard source_zone=" << sync_env->source_zone << " shard_id=" << shard_id; | |
1119 | status_oid = RGWDataSyncStatusManager::shard_obj_name(sync_env->source_zone, shard_id); | |
1120 | error_oid = status_oid + ".retry"; | |
1121 | ||
1122 | logger.init(sync_env, "DataShard", status_oid); | |
1123 | } | |
1124 | ||
1125 | ~RGWDataSyncShardCR() override { | |
1126 | delete marker_tracker; | |
1127 | if (lease_cr) { | |
1128 | lease_cr->abort(); | |
7c673cae FG |
1129 | } |
1130 | if (error_repo) { | |
1131 | error_repo->put(); | |
1132 | } | |
1133 | } | |
1134 | ||
1135 | void append_modified_shards(set<string>& keys) { | |
1136 | Mutex::Locker l(inc_lock); | |
1137 | modified_shards.insert(keys.begin(), keys.end()); | |
1138 | } | |
1139 | ||
1140 | void set_marker_tracker(RGWDataSyncShardMarkerTrack *mt) { | |
1141 | delete marker_tracker; | |
1142 | marker_tracker = mt; | |
1143 | } | |
1144 | ||
1145 | int operate() override { | |
1146 | int r; | |
1147 | while (true) { | |
1148 | switch (sync_marker.state) { | |
1149 | case rgw_data_sync_marker::FullSync: | |
1150 | r = full_sync(); | |
1151 | if (r < 0) { | |
1152 | ldout(cct, 10) << "sync: full_sync: shard_id=" << shard_id << " r=" << r << dendl; | |
1153 | return set_cr_error(r); | |
1154 | } | |
1155 | return 0; | |
1156 | case rgw_data_sync_marker::IncrementalSync: | |
1157 | r = incremental_sync(); | |
1158 | if (r < 0) { | |
1159 | ldout(cct, 10) << "sync: incremental_sync: shard_id=" << shard_id << " r=" << r << dendl; | |
1160 | return set_cr_error(r); | |
1161 | } | |
1162 | return 0; | |
1163 | default: | |
1164 | return set_cr_error(-EIO); | |
1165 | } | |
1166 | } | |
1167 | return 0; | |
1168 | } | |
1169 | ||
1170 | void init_lease_cr() { | |
1171 | set_status("acquiring sync lock"); | |
1172 | uint32_t lock_duration = cct->_conf->rgw_sync_lease_period; | |
1173 | string lock_name = "sync_lock"; | |
1174 | if (lease_cr) { | |
1175 | lease_cr->abort(); | |
7c673cae FG |
1176 | } |
1177 | RGWRados *store = sync_env->store; | |
31f18b77 FG |
1178 | lease_cr.reset(new RGWContinuousLeaseCR(sync_env->async_rados, store, |
1179 | rgw_raw_obj(store->get_zone_params().log_pool, status_oid), | |
1180 | lock_name, lock_duration, this)); | |
1181 | lease_stack.reset(spawn(lease_cr.get(), false)); | |
7c673cae FG |
1182 | } |
1183 | ||
1184 | int full_sync() { | |
1185 | #define OMAP_GET_MAX_ENTRIES 100 | |
1186 | int max_entries = OMAP_GET_MAX_ENTRIES; | |
1187 | reenter(&full_cr) { | |
1188 | yield init_lease_cr(); | |
1189 | while (!lease_cr->is_locked()) { | |
1190 | if (lease_cr->is_done()) { | |
1191 | ldout(cct, 5) << "lease cr failed, done early " << dendl; | |
1192 | set_status("lease lock failed, early abort"); | |
1193 | return set_cr_error(lease_cr->get_ret_status()); | |
1194 | } | |
1195 | set_sleeping(true); | |
1196 | yield; | |
1197 | } | |
1198 | logger.log("full sync"); | |
1199 | oid = full_data_sync_index_shard_oid(sync_env->source_zone, shard_id); | |
1200 | set_marker_tracker(new RGWDataSyncShardMarkerTrack(sync_env, status_oid, sync_marker)); | |
1201 | total_entries = sync_marker.pos; | |
1202 | do { | |
1203 | yield call(new RGWRadosGetOmapKeysCR(sync_env->store, rgw_raw_obj(pool, oid), sync_marker.marker, &entries, max_entries)); | |
1204 | if (retcode < 0) { | |
1205 | ldout(sync_env->cct, 0) << "ERROR: " << __func__ << "(): RGWRadosGetOmapKeysCR() returned ret=" << retcode << dendl; | |
1206 | lease_cr->go_down(); | |
1207 | drain_all(); | |
1208 | return set_cr_error(retcode); | |
1209 | } | |
1210 | iter = entries.begin(); | |
1211 | for (; iter != entries.end(); ++iter) { | |
1212 | ldout(sync_env->cct, 20) << __func__ << ": full sync: " << iter->first << dendl; | |
1213 | total_entries++; | |
1214 | if (!marker_tracker->start(iter->first, total_entries, real_time())) { | |
1215 | ldout(sync_env->cct, 0) << "ERROR: cannot start syncing " << iter->first << ". Duplicate entry?" << dendl; | |
1216 | } else { | |
1217 | // fetch remote and write locally | |
1218 | yield spawn(new RGWDataSyncSingleEntryCR(sync_env, iter->first, iter->first, marker_tracker, error_repo, false), false); | |
1219 | if (retcode < 0) { | |
1220 | lease_cr->go_down(); | |
1221 | drain_all(); | |
1222 | return set_cr_error(retcode); | |
1223 | } | |
1224 | } | |
1225 | sync_marker.marker = iter->first; | |
1226 | } | |
1227 | } while ((int)entries.size() == max_entries); | |
1228 | ||
1229 | lease_cr->go_down(); | |
1230 | drain_all(); | |
1231 | ||
1232 | yield { | |
1233 | /* update marker to reflect we're done with full sync */ | |
1234 | sync_marker.state = rgw_data_sync_marker::IncrementalSync; | |
1235 | sync_marker.marker = sync_marker.next_step_marker; | |
1236 | sync_marker.next_step_marker.clear(); | |
1237 | RGWRados *store = sync_env->store; | |
1238 | call(new RGWSimpleRadosWriteCR<rgw_data_sync_marker>(sync_env->async_rados, store, | |
1239 | rgw_raw_obj(store->get_zone_params().log_pool, status_oid), | |
1240 | sync_marker)); | |
1241 | } | |
1242 | if (retcode < 0) { | |
1243 | ldout(sync_env->cct, 0) << "ERROR: failed to set sync marker: retcode=" << retcode << dendl; | |
1244 | lease_cr->go_down(); | |
1245 | return set_cr_error(retcode); | |
1246 | } | |
1247 | } | |
1248 | return 0; | |
1249 | } | |
1250 | ||
1251 | int incremental_sync() { | |
1252 | reenter(&incremental_cr) { | |
1253 | yield init_lease_cr(); | |
1254 | while (!lease_cr->is_locked()) { | |
1255 | if (lease_cr->is_done()) { | |
1256 | ldout(cct, 5) << "lease cr failed, done early " << dendl; | |
1257 | set_status("lease lock failed, early abort"); | |
1258 | return set_cr_error(lease_cr->get_ret_status()); | |
1259 | } | |
1260 | set_sleeping(true); | |
1261 | yield; | |
1262 | } | |
1263 | set_status("lease acquired"); | |
1264 | error_repo = new RGWOmapAppend(sync_env->async_rados, sync_env->store, | |
1265 | rgw_raw_obj(pool, error_oid), | |
1266 | 1 /* no buffer */); | |
1267 | error_repo->get(); | |
1268 | spawn(error_repo, false); | |
1269 | logger.log("inc sync"); | |
1270 | set_marker_tracker(new RGWDataSyncShardMarkerTrack(sync_env, status_oid, sync_marker)); | |
1271 | do { | |
1272 | current_modified.clear(); | |
1273 | inc_lock.Lock(); | |
1274 | current_modified.swap(modified_shards); | |
1275 | inc_lock.Unlock(); | |
1276 | ||
1277 | /* process out of band updates */ | |
1278 | for (modified_iter = current_modified.begin(); modified_iter != current_modified.end(); ++modified_iter) { | |
1279 | yield { | |
1280 | ldout(sync_env->cct, 20) << __func__ << "(): async update notification: " << *modified_iter << dendl; | |
1281 | spawn(new RGWDataSyncSingleEntryCR(sync_env, *modified_iter, string(), marker_tracker, error_repo, false), false); | |
1282 | } | |
1283 | } | |
1284 | ||
1285 | /* process bucket shards that previously failed */ | |
1286 | yield call(new RGWRadosGetOmapKeysCR(sync_env->store, rgw_raw_obj(pool, error_oid), | |
1287 | error_marker, &error_entries, | |
1288 | max_error_entries)); | |
1289 | ldout(sync_env->cct, 20) << __func__ << "(): read error repo, got " << error_entries.size() << " entries" << dendl; | |
1290 | iter = error_entries.begin(); | |
1291 | for (; iter != error_entries.end(); ++iter) { | |
1292 | ldout(sync_env->cct, 20) << __func__ << "(): handle error entry: " << iter->first << dendl; | |
1293 | spawn(new RGWDataSyncSingleEntryCR(sync_env, iter->first, iter->first, nullptr /* no marker tracker */, error_repo, true), false); | |
1294 | error_marker = iter->first; | |
1295 | } | |
1296 | if ((int)error_entries.size() != max_error_entries) { | |
1297 | if (error_marker.empty() && error_entries.empty()) { | |
1298 | /* the retry repo is empty, we back off a bit before calling it again */ | |
1299 | retry_backoff_secs *= 2; | |
1300 | if (retry_backoff_secs > RETRY_BACKOFF_SECS_MAX) { | |
1301 | retry_backoff_secs = RETRY_BACKOFF_SECS_MAX; | |
1302 | } | |
1303 | } else { | |
1304 | retry_backoff_secs = RETRY_BACKOFF_SECS_DEFAULT; | |
1305 | } | |
1306 | error_retry_time = ceph::real_clock::now() + make_timespan(retry_backoff_secs); | |
1307 | error_marker.clear(); | |
1308 | } | |
1309 | ||
1310 | ||
1311 | yield call(new RGWReadRemoteDataLogShardInfoCR(sync_env, shard_id, &shard_info)); | |
1312 | if (retcode < 0) { | |
1313 | ldout(sync_env->cct, 0) << "ERROR: failed to fetch remote data log info: ret=" << retcode << dendl; | |
1314 | stop_spawned_services(); | |
1315 | drain_all(); | |
1316 | return set_cr_error(retcode); | |
1317 | } | |
1318 | datalog_marker = shard_info.marker; | |
1319 | #define INCREMENTAL_MAX_ENTRIES 100 | |
1320 | ldout(sync_env->cct, 20) << __func__ << ":" << __LINE__ << ": shard_id=" << shard_id << " datalog_marker=" << datalog_marker << " sync_marker.marker=" << sync_marker.marker << dendl; | |
1321 | if (datalog_marker > sync_marker.marker) { | |
1322 | spawned_keys.clear(); | |
1323 | yield call(new RGWReadRemoteDataLogShardCR(sync_env, shard_id, &sync_marker.marker, &log_entries, &truncated)); | |
1324 | if (retcode < 0) { | |
1325 | ldout(sync_env->cct, 0) << "ERROR: failed to read remote data log info: ret=" << retcode << dendl; | |
1326 | stop_spawned_services(); | |
1327 | drain_all(); | |
1328 | return set_cr_error(retcode); | |
1329 | } | |
1330 | for (log_iter = log_entries.begin(); log_iter != log_entries.end(); ++log_iter) { | |
1331 | ldout(sync_env->cct, 20) << __func__ << ":" << __LINE__ << ": shard_id=" << shard_id << " log_entry: " << log_iter->log_id << ":" << log_iter->log_timestamp << ":" << log_iter->entry.key << dendl; | |
1332 | if (!marker_tracker->index_key_to_marker(log_iter->entry.key, log_iter->log_id)) { | |
1333 | ldout(sync_env->cct, 20) << __func__ << ": skipping sync of entry: " << log_iter->log_id << ":" << log_iter->entry.key << " sync already in progress for bucket shard" << dendl; | |
1334 | marker_tracker->try_update_high_marker(log_iter->log_id, 0, log_iter->log_timestamp); | |
1335 | continue; | |
1336 | } | |
1337 | if (!marker_tracker->start(log_iter->log_id, 0, log_iter->log_timestamp)) { | |
1338 | ldout(sync_env->cct, 0) << "ERROR: cannot start syncing " << log_iter->log_id << ". Duplicate entry?" << dendl; | |
1339 | } else { | |
1340 | /* | |
1341 | * don't spawn the same key more than once. We can do that as long as we don't yield | |
1342 | */ | |
1343 | if (spawned_keys.find(log_iter->entry.key) == spawned_keys.end()) { | |
1344 | spawned_keys.insert(log_iter->entry.key); | |
1345 | spawn(new RGWDataSyncSingleEntryCR(sync_env, log_iter->entry.key, log_iter->log_id, marker_tracker, error_repo, false), false); | |
1346 | if (retcode < 0) { | |
1347 | stop_spawned_services(); | |
1348 | drain_all(); | |
1349 | return set_cr_error(retcode); | |
1350 | } | |
1351 | } | |
1352 | } | |
1353 | } | |
1354 | while ((int)num_spawned() > spawn_window) { | |
1355 | set_status() << "num_spawned() > spawn_window"; | |
1356 | yield wait_for_child(); | |
1357 | int ret; | |
31f18b77 | 1358 | while (collect(&ret, lease_stack.get())) { |
7c673cae FG |
1359 | if (ret < 0) { |
1360 | ldout(sync_env->cct, 0) << "ERROR: a sync operation returned error" << dendl; | |
1361 | /* we have reported this error */ | |
1362 | } | |
1363 | /* not waiting for child here */ | |
1364 | } | |
1365 | } | |
1366 | } | |
1367 | ldout(sync_env->cct, 20) << __func__ << ":" << __LINE__ << ": shard_id=" << shard_id << " datalog_marker=" << datalog_marker << " sync_marker.marker=" << sync_marker.marker << dendl; | |
1368 | if (datalog_marker == sync_marker.marker) { | |
1369 | #define INCREMENTAL_INTERVAL 20 | |
1370 | yield wait(utime_t(INCREMENTAL_INTERVAL, 0)); | |
1371 | } | |
1372 | } while (true); | |
1373 | } | |
1374 | return 0; | |
1375 | } | |
1376 | void stop_spawned_services() { | |
1377 | lease_cr->go_down(); | |
1378 | if (error_repo) { | |
1379 | error_repo->finish(); | |
1380 | error_repo->put(); | |
1381 | error_repo = NULL; | |
1382 | } | |
1383 | } | |
1384 | }; | |
1385 | ||
1386 | class RGWDataSyncShardControlCR : public RGWBackoffControlCR { | |
1387 | RGWDataSyncEnv *sync_env; | |
1388 | ||
1389 | rgw_pool pool; | |
1390 | ||
1391 | uint32_t shard_id; | |
1392 | rgw_data_sync_marker sync_marker; | |
1393 | ||
1394 | public: | |
1395 | RGWDataSyncShardControlCR(RGWDataSyncEnv *_sync_env, rgw_pool& _pool, | |
1396 | uint32_t _shard_id, rgw_data_sync_marker& _marker) : RGWBackoffControlCR(_sync_env->cct, false), | |
1397 | sync_env(_sync_env), | |
1398 | pool(_pool), | |
1399 | shard_id(_shard_id), | |
1400 | sync_marker(_marker) { | |
1401 | } | |
1402 | ||
1403 | RGWCoroutine *alloc_cr() override { | |
1404 | return new RGWDataSyncShardCR(sync_env, pool, shard_id, sync_marker, backoff_ptr()); | |
1405 | } | |
1406 | ||
1407 | RGWCoroutine *alloc_finisher_cr() override { | |
1408 | RGWRados *store = sync_env->store; | |
1409 | return new RGWSimpleRadosReadCR<rgw_data_sync_marker>(sync_env->async_rados, store, | |
1410 | rgw_raw_obj(store->get_zone_params().log_pool, RGWDataSyncStatusManager::shard_obj_name(sync_env->source_zone, shard_id)), | |
1411 | &sync_marker); | |
1412 | } | |
1413 | ||
1414 | void append_modified_shards(set<string>& keys) { | |
1415 | Mutex::Locker l(cr_lock()); | |
1416 | ||
1417 | RGWDataSyncShardCR *cr = static_cast<RGWDataSyncShardCR *>(get_cr()); | |
1418 | if (!cr) { | |
1419 | return; | |
1420 | } | |
1421 | ||
1422 | cr->append_modified_shards(keys); | |
1423 | } | |
1424 | }; | |
1425 | ||
1426 | class RGWDataSyncCR : public RGWCoroutine { | |
1427 | RGWDataSyncEnv *sync_env; | |
1428 | uint32_t num_shards; | |
1429 | ||
1430 | rgw_data_sync_status sync_status; | |
1431 | ||
1432 | RGWDataSyncShardMarkerTrack *marker_tracker; | |
1433 | ||
1434 | Mutex shard_crs_lock; | |
1435 | map<int, RGWDataSyncShardControlCR *> shard_crs; | |
1436 | ||
1437 | bool *reset_backoff; | |
1438 | ||
1439 | RGWDataSyncDebugLogger logger; | |
31f18b77 FG |
1440 | |
1441 | RGWDataSyncModule *data_sync_module{nullptr}; | |
7c673cae FG |
1442 | public: |
1443 | RGWDataSyncCR(RGWDataSyncEnv *_sync_env, uint32_t _num_shards, bool *_reset_backoff) : RGWCoroutine(_sync_env->cct), | |
1444 | sync_env(_sync_env), | |
1445 | num_shards(_num_shards), | |
1446 | marker_tracker(NULL), | |
1447 | shard_crs_lock("RGWDataSyncCR::shard_crs_lock"), | |
1448 | reset_backoff(_reset_backoff), logger(sync_env, "Data", "all") { | |
31f18b77 | 1449 | |
7c673cae FG |
1450 | } |
1451 | ||
1452 | ~RGWDataSyncCR() override { | |
1453 | for (auto iter : shard_crs) { | |
1454 | iter.second->put(); | |
1455 | } | |
1456 | } | |
1457 | ||
1458 | int operate() override { | |
1459 | reenter(this) { | |
1460 | ||
1461 | /* read sync status */ | |
1462 | yield call(new RGWReadDataSyncStatusCoroutine(sync_env, &sync_status)); | |
1463 | ||
31f18b77 FG |
1464 | data_sync_module = sync_env->sync_module->get_data_handler(); |
1465 | ||
7c673cae FG |
1466 | if (retcode == -ENOENT) { |
1467 | sync_status.sync_info.num_shards = num_shards; | |
1468 | } else if (retcode < 0 && retcode != -ENOENT) { | |
1469 | ldout(sync_env->cct, 0) << "ERROR: failed to fetch sync status, retcode=" << retcode << dendl; | |
1470 | return set_cr_error(retcode); | |
1471 | } | |
1472 | ||
1473 | /* state: init status */ | |
1474 | if ((rgw_data_sync_info::SyncState)sync_status.sync_info.state == rgw_data_sync_info::StateInit) { | |
1475 | ldout(sync_env->cct, 20) << __func__ << "(): init" << dendl; | |
31f18b77 FG |
1476 | uint64_t instance_id; |
1477 | get_random_bytes((char *)&instance_id, sizeof(instance_id)); | |
1478 | yield call(new RGWInitDataSyncStatusCoroutine(sync_env, num_shards, instance_id, &sync_status)); | |
7c673cae FG |
1479 | if (retcode < 0) { |
1480 | ldout(sync_env->cct, 0) << "ERROR: failed to init sync, retcode=" << retcode << dendl; | |
1481 | return set_cr_error(retcode); | |
1482 | } | |
1483 | // sets state = StateBuildingFullSyncMaps | |
1484 | ||
1485 | *reset_backoff = true; | |
1486 | } | |
1487 | ||
31f18b77 FG |
1488 | data_sync_module->init(sync_env, sync_status.sync_info.instance_id); |
1489 | ||
7c673cae | 1490 | if ((rgw_data_sync_info::SyncState)sync_status.sync_info.state == rgw_data_sync_info::StateBuildingFullSyncMaps) { |
31f18b77 FG |
1491 | /* call sync module init here */ |
1492 | yield call(data_sync_module->init_sync(sync_env)); | |
1493 | if (retcode < 0) { | |
1494 | ldout(sync_env->cct, 0) << "ERROR: sync module init_sync() failed, retcode=" << retcode << dendl; | |
1495 | return set_cr_error(retcode); | |
1496 | } | |
7c673cae FG |
1497 | /* state: building full sync maps */ |
1498 | ldout(sync_env->cct, 20) << __func__ << "(): building full sync maps" << dendl; | |
1499 | yield call(new RGWListBucketIndexesCR(sync_env, &sync_status)); | |
1500 | if (retcode < 0) { | |
1501 | ldout(sync_env->cct, 0) << "ERROR: failed to build full sync maps, retcode=" << retcode << dendl; | |
1502 | return set_cr_error(retcode); | |
1503 | } | |
1504 | sync_status.sync_info.state = rgw_data_sync_info::StateSync; | |
1505 | ||
1506 | /* update new state */ | |
1507 | yield call(set_sync_info_cr()); | |
1508 | if (retcode < 0) { | |
1509 | ldout(sync_env->cct, 0) << "ERROR: failed to write sync status, retcode=" << retcode << dendl; | |
1510 | return set_cr_error(retcode); | |
1511 | } | |
1512 | ||
1513 | *reset_backoff = true; | |
1514 | } | |
1515 | ||
1516 | yield { | |
1517 | if ((rgw_data_sync_info::SyncState)sync_status.sync_info.state == rgw_data_sync_info::StateSync) { | |
1518 | for (map<uint32_t, rgw_data_sync_marker>::iterator iter = sync_status.sync_markers.begin(); | |
1519 | iter != sync_status.sync_markers.end(); ++iter) { | |
1520 | RGWDataSyncShardControlCR *cr = new RGWDataSyncShardControlCR(sync_env, sync_env->store->get_zone_params().log_pool, | |
1521 | iter->first, iter->second); | |
1522 | cr->get(); | |
1523 | shard_crs_lock.Lock(); | |
1524 | shard_crs[iter->first] = cr; | |
1525 | shard_crs_lock.Unlock(); | |
1526 | spawn(cr, true); | |
1527 | } | |
1528 | } | |
1529 | } | |
1530 | ||
1531 | return set_cr_done(); | |
1532 | } | |
1533 | return 0; | |
1534 | } | |
1535 | ||
1536 | RGWCoroutine *set_sync_info_cr() { | |
1537 | RGWRados *store = sync_env->store; | |
1538 | return new RGWSimpleRadosWriteCR<rgw_data_sync_info>(sync_env->async_rados, store, | |
1539 | rgw_raw_obj(store->get_zone_params().log_pool, RGWDataSyncStatusManager::sync_status_oid(sync_env->source_zone)), | |
1540 | sync_status.sync_info); | |
1541 | } | |
1542 | ||
1543 | void wakeup(int shard_id, set<string>& keys) { | |
1544 | Mutex::Locker l(shard_crs_lock); | |
1545 | map<int, RGWDataSyncShardControlCR *>::iterator iter = shard_crs.find(shard_id); | |
1546 | if (iter == shard_crs.end()) { | |
1547 | return; | |
1548 | } | |
1549 | iter->second->append_modified_shards(keys); | |
1550 | iter->second->wakeup(); | |
1551 | } | |
1552 | }; | |
1553 | ||
1554 | class RGWDefaultDataSyncModule : public RGWDataSyncModule { | |
1555 | public: | |
1556 | RGWDefaultDataSyncModule() {} | |
1557 | ||
31f18b77 FG |
1558 | RGWCoroutine *sync_object(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, uint64_t versioned_epoch, rgw_zone_set *zones_trace) override; |
1559 | RGWCoroutine *remove_object(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, real_time& mtime, bool versioned, uint64_t versioned_epoch, rgw_zone_set *zones_trace) override; | |
7c673cae | 1560 | RGWCoroutine *create_delete_marker(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, real_time& mtime, |
31f18b77 | 1561 | rgw_bucket_entry_owner& owner, bool versioned, uint64_t versioned_epoch, rgw_zone_set *zones_trace) override; |
7c673cae FG |
1562 | }; |
1563 | ||
1564 | class RGWDefaultSyncModuleInstance : public RGWSyncModuleInstance { | |
1565 | RGWDefaultDataSyncModule data_handler; | |
1566 | public: | |
1567 | RGWDefaultSyncModuleInstance() {} | |
1568 | RGWDataSyncModule *get_data_handler() override { | |
1569 | return &data_handler; | |
1570 | } | |
1571 | }; | |
1572 | ||
31f18b77 | 1573 | int RGWDefaultSyncModule::create_instance(CephContext *cct, map<string, string, ltstr_nocase>& config, RGWSyncModuleInstanceRef *instance) |
7c673cae FG |
1574 | { |
1575 | instance->reset(new RGWDefaultSyncModuleInstance()); | |
1576 | return 0; | |
1577 | } | |
1578 | ||
31f18b77 | 1579 | RGWCoroutine *RGWDefaultDataSyncModule::sync_object(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, uint64_t versioned_epoch, rgw_zone_set *zones_trace) |
7c673cae FG |
1580 | { |
1581 | return new RGWFetchRemoteObjCR(sync_env->async_rados, sync_env->store, sync_env->source_zone, bucket_info, | |
1582 | key, versioned_epoch, | |
31f18b77 | 1583 | true, zones_trace); |
7c673cae FG |
1584 | } |
1585 | ||
1586 | RGWCoroutine *RGWDefaultDataSyncModule::remove_object(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, | |
31f18b77 | 1587 | real_time& mtime, bool versioned, uint64_t versioned_epoch, rgw_zone_set *zones_trace) |
7c673cae FG |
1588 | { |
1589 | return new RGWRemoveObjCR(sync_env->async_rados, sync_env->store, sync_env->source_zone, | |
1590 | bucket_info, key, versioned, versioned_epoch, | |
31f18b77 | 1591 | NULL, NULL, false, &mtime, zones_trace); |
7c673cae FG |
1592 | } |
1593 | ||
1594 | RGWCoroutine *RGWDefaultDataSyncModule::create_delete_marker(RGWDataSyncEnv *sync_env, RGWBucketInfo& bucket_info, rgw_obj_key& key, real_time& mtime, | |
31f18b77 | 1595 | rgw_bucket_entry_owner& owner, bool versioned, uint64_t versioned_epoch, rgw_zone_set *zones_trace) |
7c673cae FG |
1596 | { |
1597 | return new RGWRemoveObjCR(sync_env->async_rados, sync_env->store, sync_env->source_zone, | |
1598 | bucket_info, key, versioned, versioned_epoch, | |
31f18b77 | 1599 | &owner.id, &owner.display_name, true, &mtime, zones_trace); |
7c673cae FG |
1600 | } |
1601 | ||
1602 | class RGWDataSyncControlCR : public RGWBackoffControlCR | |
1603 | { | |
1604 | RGWDataSyncEnv *sync_env; | |
1605 | uint32_t num_shards; | |
1606 | ||
1607 | public: | |
1608 | RGWDataSyncControlCR(RGWDataSyncEnv *_sync_env, uint32_t _num_shards) : RGWBackoffControlCR(_sync_env->cct, true), | |
1609 | sync_env(_sync_env), num_shards(_num_shards) { | |
1610 | } | |
1611 | ||
1612 | RGWCoroutine *alloc_cr() override { | |
1613 | return new RGWDataSyncCR(sync_env, num_shards, backoff_ptr()); | |
1614 | } | |
1615 | ||
1616 | void wakeup(int shard_id, set<string>& keys) { | |
1617 | Mutex& m = cr_lock(); | |
1618 | ||
1619 | m.Lock(); | |
1620 | RGWDataSyncCR *cr = static_cast<RGWDataSyncCR *>(get_cr()); | |
1621 | if (!cr) { | |
1622 | m.Unlock(); | |
1623 | return; | |
1624 | } | |
1625 | ||
1626 | cr->get(); | |
1627 | m.Unlock(); | |
1628 | ||
1629 | if (cr) { | |
1630 | cr->wakeup(shard_id, keys); | |
1631 | } | |
1632 | ||
1633 | cr->put(); | |
1634 | } | |
1635 | }; | |
1636 | ||
1637 | void RGWRemoteDataLog::wakeup(int shard_id, set<string>& keys) { | |
1638 | RWLock::RLocker rl(lock); | |
1639 | if (!data_sync_cr) { | |
1640 | return; | |
1641 | } | |
1642 | data_sync_cr->wakeup(shard_id, keys); | |
1643 | } | |
1644 | ||
1645 | int RGWRemoteDataLog::run_sync(int num_shards) | |
1646 | { | |
1647 | lock.get_write(); | |
1648 | data_sync_cr = new RGWDataSyncControlCR(&sync_env, num_shards); | |
1649 | data_sync_cr->get(); // run() will drop a ref, so take another | |
1650 | lock.unlock(); | |
1651 | ||
1652 | int r = run(data_sync_cr); | |
1653 | ||
1654 | lock.get_write(); | |
1655 | data_sync_cr->put(); | |
1656 | data_sync_cr = NULL; | |
1657 | lock.unlock(); | |
1658 | ||
1659 | if (r < 0) { | |
1660 | ldout(store->ctx(), 0) << "ERROR: failed to run sync" << dendl; | |
1661 | return r; | |
1662 | } | |
1663 | return 0; | |
1664 | } | |
1665 | ||
1666 | int RGWDataSyncStatusManager::init() | |
1667 | { | |
1668 | auto zone_def_iter = store->zone_by_id.find(source_zone); | |
1669 | if (zone_def_iter == store->zone_by_id.end()) { | |
1670 | ldout(store->ctx(), 0) << "ERROR: failed to find zone config info for zone=" << source_zone << dendl; | |
1671 | return -EIO; | |
1672 | } | |
1673 | ||
1674 | auto& zone_def = zone_def_iter->second; | |
1675 | ||
1676 | if (!store->get_sync_modules_manager()->supports_data_export(zone_def.tier_type)) { | |
1677 | return -ENOTSUP; | |
1678 | } | |
1679 | ||
1680 | RGWZoneParams& zone_params = store->get_zone_params(); | |
1681 | ||
1682 | sync_module = store->get_sync_module(); | |
1683 | ||
1684 | conn = store->get_zone_conn_by_id(source_zone); | |
1685 | if (!conn) { | |
1686 | ldout(store->ctx(), 0) << "connection object to zone " << source_zone << " does not exist" << dendl; | |
1687 | return -EINVAL; | |
1688 | } | |
1689 | ||
1690 | error_logger = new RGWSyncErrorLogger(store, RGW_SYNC_ERROR_LOG_SHARD_PREFIX, ERROR_LOGGER_SHARDS); | |
1691 | ||
1692 | int r = source_log.init(source_zone, conn, error_logger, sync_module); | |
1693 | if (r < 0) { | |
1694 | lderr(store->ctx()) << "ERROR: failed to init remote log, r=" << r << dendl; | |
1695 | finalize(); | |
1696 | return r; | |
1697 | } | |
1698 | ||
1699 | rgw_datalog_info datalog_info; | |
1700 | r = source_log.read_log_info(&datalog_info); | |
1701 | if (r < 0) { | |
1702 | ldout(store->ctx(), 5) << "ERROR: master.read_log_info() returned r=" << r << dendl; | |
1703 | finalize(); | |
1704 | return r; | |
1705 | } | |
1706 | ||
1707 | num_shards = datalog_info.num_shards; | |
1708 | ||
1709 | for (int i = 0; i < num_shards; i++) { | |
1710 | shard_objs[i] = rgw_raw_obj(zone_params.log_pool, shard_obj_name(source_zone, i)); | |
1711 | } | |
1712 | ||
1713 | return 0; | |
1714 | } | |
1715 | ||
1716 | void RGWDataSyncStatusManager::finalize() | |
1717 | { | |
1718 | delete error_logger; | |
1719 | error_logger = nullptr; | |
1720 | } | |
1721 | ||
1722 | string RGWDataSyncStatusManager::sync_status_oid(const string& source_zone) | |
1723 | { | |
1724 | char buf[datalog_sync_status_oid_prefix.size() + source_zone.size() + 16]; | |
1725 | snprintf(buf, sizeof(buf), "%s.%s", datalog_sync_status_oid_prefix.c_str(), source_zone.c_str()); | |
1726 | ||
1727 | return string(buf); | |
1728 | } | |
1729 | ||
1730 | string RGWDataSyncStatusManager::shard_obj_name(const string& source_zone, int shard_id) | |
1731 | { | |
1732 | char buf[datalog_sync_status_shard_prefix.size() + source_zone.size() + 16]; | |
1733 | snprintf(buf, sizeof(buf), "%s.%s.%d", datalog_sync_status_shard_prefix.c_str(), source_zone.c_str(), shard_id); | |
1734 | ||
1735 | return string(buf); | |
1736 | } | |
1737 | ||
1738 | int RGWRemoteBucketLog::init(const string& _source_zone, RGWRESTConn *_conn, | |
1739 | const rgw_bucket& bucket, int shard_id, | |
1740 | RGWSyncErrorLogger *_error_logger, | |
1741 | RGWSyncModuleInstanceRef& _sync_module) | |
1742 | { | |
1743 | conn = _conn; | |
1744 | source_zone = _source_zone; | |
1745 | bs.bucket = bucket; | |
1746 | bs.shard_id = shard_id; | |
1747 | ||
1748 | sync_env.init(store->ctx(), store, conn, async_rados, http_manager, _error_logger, source_zone, _sync_module); | |
1749 | ||
1750 | return 0; | |
1751 | } | |
1752 | ||
1753 | struct bucket_index_marker_info { | |
1754 | string bucket_ver; | |
1755 | string master_ver; | |
1756 | string max_marker; | |
1757 | ||
1758 | void decode_json(JSONObj *obj) { | |
1759 | JSONDecoder::decode_json("bucket_ver", bucket_ver, obj); | |
1760 | JSONDecoder::decode_json("master_ver", master_ver, obj); | |
1761 | JSONDecoder::decode_json("max_marker", max_marker, obj); | |
1762 | } | |
1763 | }; | |
1764 | ||
1765 | class RGWReadRemoteBucketIndexLogInfoCR : public RGWCoroutine { | |
1766 | RGWDataSyncEnv *sync_env; | |
1767 | const string instance_key; | |
1768 | ||
1769 | bucket_index_marker_info *info; | |
1770 | ||
1771 | public: | |
1772 | RGWReadRemoteBucketIndexLogInfoCR(RGWDataSyncEnv *_sync_env, | |
1773 | const rgw_bucket_shard& bs, | |
1774 | bucket_index_marker_info *_info) | |
1775 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), | |
1776 | instance_key(bs.get_key()), info(_info) {} | |
1777 | ||
1778 | int operate() override { | |
1779 | reenter(this) { | |
1780 | yield { | |
1781 | rgw_http_param_pair pairs[] = { { "type" , "bucket-index" }, | |
1782 | { "bucket-instance", instance_key.c_str() }, | |
1783 | { "info" , NULL }, | |
1784 | { NULL, NULL } }; | |
1785 | ||
1786 | string p = "/admin/log/"; | |
1787 | call(new RGWReadRESTResourceCR<bucket_index_marker_info>(sync_env->cct, sync_env->conn, sync_env->http_manager, p, pairs, info)); | |
1788 | } | |
1789 | if (retcode < 0) { | |
1790 | return set_cr_error(retcode); | |
1791 | } | |
1792 | return set_cr_done(); | |
1793 | } | |
1794 | return 0; | |
1795 | } | |
1796 | }; | |
1797 | ||
1798 | class RGWInitBucketShardSyncStatusCoroutine : public RGWCoroutine { | |
1799 | RGWDataSyncEnv *sync_env; | |
1800 | ||
1801 | rgw_bucket_shard bs; | |
1802 | const string sync_status_oid; | |
1803 | ||
1804 | rgw_bucket_shard_sync_info& status; | |
1805 | ||
1806 | bucket_index_marker_info info; | |
1807 | public: | |
1808 | RGWInitBucketShardSyncStatusCoroutine(RGWDataSyncEnv *_sync_env, | |
1809 | const rgw_bucket_shard& bs, | |
1810 | rgw_bucket_shard_sync_info& _status) | |
1811 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), bs(bs), | |
1812 | sync_status_oid(RGWBucketSyncStatusManager::status_oid(sync_env->source_zone, bs)), | |
1813 | status(_status) | |
1814 | {} | |
1815 | ||
1816 | int operate() override { | |
1817 | reenter(this) { | |
1818 | /* fetch current position in logs */ | |
1819 | yield call(new RGWReadRemoteBucketIndexLogInfoCR(sync_env, bs, &info)); | |
1820 | if (retcode < 0 && retcode != -ENOENT) { | |
1821 | ldout(cct, 0) << "ERROR: failed to fetch bucket index status" << dendl; | |
1822 | return set_cr_error(retcode); | |
1823 | } | |
1824 | yield { | |
1825 | status.state = rgw_bucket_shard_sync_info::StateFullSync; | |
1826 | status.inc_marker.position = info.max_marker; | |
1827 | map<string, bufferlist> attrs; | |
1828 | status.encode_all_attrs(attrs); | |
1829 | auto store = sync_env->store; | |
1830 | call(new RGWSimpleRadosWriteAttrsCR(sync_env->async_rados, store, | |
1831 | rgw_raw_obj(store->get_zone_params().log_pool, sync_status_oid), | |
1832 | attrs)); | |
1833 | } | |
1834 | return set_cr_done(); | |
1835 | } | |
1836 | return 0; | |
1837 | } | |
1838 | }; | |
1839 | ||
1840 | RGWCoroutine *RGWRemoteBucketLog::init_sync_status_cr() | |
1841 | { | |
1842 | return new RGWInitBucketShardSyncStatusCoroutine(&sync_env, bs, init_status); | |
1843 | } | |
1844 | ||
1845 | template <class T> | |
1846 | static void decode_attr(CephContext *cct, map<string, bufferlist>& attrs, const string& attr_name, T *val) | |
1847 | { | |
1848 | map<string, bufferlist>::iterator iter = attrs.find(attr_name); | |
1849 | if (iter == attrs.end()) { | |
1850 | *val = T(); | |
1851 | return; | |
1852 | } | |
1853 | ||
1854 | bufferlist::iterator biter = iter->second.begin(); | |
1855 | try { | |
1856 | ::decode(*val, biter); | |
1857 | } catch (buffer::error& err) { | |
1858 | ldout(cct, 0) << "ERROR: failed to decode attribute: " << attr_name << dendl; | |
1859 | } | |
1860 | } | |
1861 | ||
1862 | void rgw_bucket_shard_sync_info::decode_from_attrs(CephContext *cct, map<string, bufferlist>& attrs) | |
1863 | { | |
1864 | decode_attr(cct, attrs, "state", &state); | |
1865 | decode_attr(cct, attrs, "full_marker", &full_marker); | |
1866 | decode_attr(cct, attrs, "inc_marker", &inc_marker); | |
1867 | } | |
1868 | ||
1869 | void rgw_bucket_shard_sync_info::encode_all_attrs(map<string, bufferlist>& attrs) | |
1870 | { | |
1871 | encode_state_attr(attrs); | |
1872 | full_marker.encode_attr(attrs); | |
1873 | inc_marker.encode_attr(attrs); | |
1874 | } | |
1875 | ||
1876 | void rgw_bucket_shard_sync_info::encode_state_attr(map<string, bufferlist>& attrs) | |
1877 | { | |
1878 | ::encode(state, attrs["state"]); | |
1879 | } | |
1880 | ||
1881 | void rgw_bucket_shard_full_sync_marker::encode_attr(map<string, bufferlist>& attrs) | |
1882 | { | |
1883 | ::encode(*this, attrs["full_marker"]); | |
1884 | } | |
1885 | ||
1886 | void rgw_bucket_shard_inc_sync_marker::encode_attr(map<string, bufferlist>& attrs) | |
1887 | { | |
1888 | ::encode(*this, attrs["inc_marker"]); | |
1889 | } | |
1890 | ||
1891 | class RGWReadBucketSyncStatusCoroutine : public RGWCoroutine { | |
1892 | RGWDataSyncEnv *sync_env; | |
1893 | string oid; | |
1894 | rgw_bucket_shard_sync_info *status; | |
1895 | ||
1896 | map<string, bufferlist> attrs; | |
1897 | public: | |
1898 | RGWReadBucketSyncStatusCoroutine(RGWDataSyncEnv *_sync_env, | |
1899 | const rgw_bucket_shard& bs, | |
1900 | rgw_bucket_shard_sync_info *_status) | |
1901 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), | |
1902 | oid(RGWBucketSyncStatusManager::status_oid(sync_env->source_zone, bs)), | |
1903 | status(_status) {} | |
1904 | int operate() override; | |
1905 | }; | |
1906 | ||
1907 | int RGWReadBucketSyncStatusCoroutine::operate() | |
1908 | { | |
1909 | reenter(this) { | |
1910 | yield call(new RGWSimpleRadosReadAttrsCR(sync_env->async_rados, sync_env->store, | |
1911 | rgw_raw_obj(sync_env->store->get_zone_params().log_pool, oid), | |
1912 | &attrs)); | |
1913 | if (retcode == -ENOENT) { | |
1914 | *status = rgw_bucket_shard_sync_info(); | |
1915 | return set_cr_done(); | |
1916 | } | |
1917 | if (retcode < 0) { | |
1918 | ldout(sync_env->cct, 0) << "ERROR: failed to call fetch bucket shard info oid=" << oid << " ret=" << retcode << dendl; | |
1919 | return set_cr_error(retcode); | |
1920 | } | |
1921 | status->decode_from_attrs(sync_env->cct, attrs); | |
1922 | return set_cr_done(); | |
1923 | } | |
1924 | return 0; | |
1925 | } | |
1926 | RGWCoroutine *RGWRemoteBucketLog::read_sync_status_cr(rgw_bucket_shard_sync_info *sync_status) | |
1927 | { | |
1928 | return new RGWReadBucketSyncStatusCoroutine(&sync_env, bs, sync_status); | |
1929 | } | |
1930 | ||
1931 | RGWBucketSyncStatusManager::~RGWBucketSyncStatusManager() { | |
1932 | for (map<int, RGWRemoteBucketLog *>::iterator iter = source_logs.begin(); iter != source_logs.end(); ++iter) { | |
1933 | delete iter->second; | |
1934 | } | |
1935 | delete error_logger; | |
1936 | } | |
1937 | ||
1938 | ||
1939 | void rgw_bucket_entry_owner::decode_json(JSONObj *obj) | |
1940 | { | |
1941 | JSONDecoder::decode_json("ID", id, obj); | |
1942 | JSONDecoder::decode_json("DisplayName", display_name, obj); | |
1943 | } | |
1944 | ||
1945 | struct bucket_list_entry { | |
1946 | bool delete_marker; | |
1947 | rgw_obj_key key; | |
1948 | bool is_latest; | |
1949 | real_time mtime; | |
1950 | string etag; | |
1951 | uint64_t size; | |
1952 | string storage_class; | |
1953 | rgw_bucket_entry_owner owner; | |
1954 | uint64_t versioned_epoch; | |
1955 | string rgw_tag; | |
1956 | ||
1957 | bucket_list_entry() : delete_marker(false), is_latest(false), size(0), versioned_epoch(0) {} | |
1958 | ||
1959 | void decode_json(JSONObj *obj) { | |
1960 | JSONDecoder::decode_json("IsDeleteMarker", delete_marker, obj); | |
1961 | JSONDecoder::decode_json("Key", key.name, obj); | |
1962 | JSONDecoder::decode_json("VersionId", key.instance, obj); | |
1963 | JSONDecoder::decode_json("IsLatest", is_latest, obj); | |
1964 | string mtime_str; | |
1965 | JSONDecoder::decode_json("RgwxMtime", mtime_str, obj); | |
1966 | ||
1967 | struct tm t; | |
1968 | uint32_t nsec; | |
1969 | if (parse_iso8601(mtime_str.c_str(), &t, &nsec)) { | |
1970 | ceph_timespec ts; | |
1971 | ts.tv_sec = (uint64_t)internal_timegm(&t); | |
1972 | ts.tv_nsec = nsec; | |
1973 | mtime = real_clock::from_ceph_timespec(ts); | |
1974 | } | |
1975 | JSONDecoder::decode_json("ETag", etag, obj); | |
1976 | JSONDecoder::decode_json("Size", size, obj); | |
1977 | JSONDecoder::decode_json("StorageClass", storage_class, obj); | |
1978 | JSONDecoder::decode_json("Owner", owner, obj); | |
1979 | JSONDecoder::decode_json("VersionedEpoch", versioned_epoch, obj); | |
1980 | JSONDecoder::decode_json("RgwxTag", rgw_tag, obj); | |
1981 | } | |
1982 | }; | |
1983 | ||
1984 | struct bucket_list_result { | |
1985 | string name; | |
1986 | string prefix; | |
1987 | string key_marker; | |
1988 | string version_id_marker; | |
1989 | int max_keys; | |
1990 | bool is_truncated; | |
1991 | list<bucket_list_entry> entries; | |
1992 | ||
1993 | bucket_list_result() : max_keys(0), is_truncated(false) {} | |
1994 | ||
1995 | void decode_json(JSONObj *obj) { | |
1996 | JSONDecoder::decode_json("Name", name, obj); | |
1997 | JSONDecoder::decode_json("Prefix", prefix, obj); | |
1998 | JSONDecoder::decode_json("KeyMarker", key_marker, obj); | |
1999 | JSONDecoder::decode_json("VersionIdMarker", version_id_marker, obj); | |
2000 | JSONDecoder::decode_json("MaxKeys", max_keys, obj); | |
2001 | JSONDecoder::decode_json("IsTruncated", is_truncated, obj); | |
2002 | JSONDecoder::decode_json("Entries", entries, obj); | |
2003 | } | |
2004 | }; | |
2005 | ||
2006 | class RGWListBucketShardCR: public RGWCoroutine { | |
2007 | RGWDataSyncEnv *sync_env; | |
2008 | const rgw_bucket_shard& bs; | |
2009 | const string instance_key; | |
2010 | rgw_obj_key marker_position; | |
2011 | ||
2012 | bucket_list_result *result; | |
2013 | ||
2014 | public: | |
2015 | RGWListBucketShardCR(RGWDataSyncEnv *_sync_env, const rgw_bucket_shard& bs, | |
2016 | rgw_obj_key& _marker_position, bucket_list_result *_result) | |
2017 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), bs(bs), | |
2018 | instance_key(bs.get_key()), marker_position(_marker_position), | |
2019 | result(_result) {} | |
2020 | ||
2021 | int operate() override { | |
2022 | reenter(this) { | |
2023 | yield { | |
2024 | rgw_http_param_pair pairs[] = { { "rgwx-bucket-instance", instance_key.c_str() }, | |
2025 | { "versions" , NULL }, | |
2026 | { "format" , "json" }, | |
2027 | { "objs-container" , "true" }, | |
2028 | { "key-marker" , marker_position.name.c_str() }, | |
2029 | { "version-id-marker" , marker_position.instance.c_str() }, | |
2030 | { NULL, NULL } }; | |
2031 | // don't include tenant in the url, it's already part of instance_key | |
2032 | string p = string("/") + bs.bucket.name; | |
2033 | call(new RGWReadRESTResourceCR<bucket_list_result>(sync_env->cct, sync_env->conn, sync_env->http_manager, p, pairs, result)); | |
2034 | } | |
2035 | if (retcode < 0) { | |
2036 | return set_cr_error(retcode); | |
2037 | } | |
2038 | return set_cr_done(); | |
2039 | } | |
2040 | return 0; | |
2041 | } | |
2042 | }; | |
2043 | ||
2044 | class RGWListBucketIndexLogCR: public RGWCoroutine { | |
2045 | RGWDataSyncEnv *sync_env; | |
2046 | const string instance_key; | |
2047 | string marker; | |
2048 | ||
2049 | list<rgw_bi_log_entry> *result; | |
2050 | ||
2051 | public: | |
2052 | RGWListBucketIndexLogCR(RGWDataSyncEnv *_sync_env, const rgw_bucket_shard& bs, | |
2053 | string& _marker, list<rgw_bi_log_entry> *_result) | |
2054 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), | |
2055 | instance_key(bs.get_key()), marker(_marker), result(_result) {} | |
2056 | ||
2057 | int operate() override { | |
2058 | reenter(this) { | |
2059 | yield { | |
2060 | rgw_http_param_pair pairs[] = { { "bucket-instance", instance_key.c_str() }, | |
2061 | { "format" , "json" }, | |
2062 | { "marker" , marker.c_str() }, | |
2063 | { "type", "bucket-index" }, | |
2064 | { NULL, NULL } }; | |
2065 | ||
2066 | call(new RGWReadRESTResourceCR<list<rgw_bi_log_entry> >(sync_env->cct, sync_env->conn, sync_env->http_manager, "/admin/log", pairs, result)); | |
2067 | } | |
2068 | if (retcode < 0) { | |
2069 | return set_cr_error(retcode); | |
2070 | } | |
2071 | return set_cr_done(); | |
2072 | } | |
2073 | return 0; | |
2074 | } | |
2075 | }; | |
2076 | ||
2077 | #define BUCKET_SYNC_UPDATE_MARKER_WINDOW 10 | |
2078 | ||
2079 | class RGWBucketFullSyncShardMarkerTrack : public RGWSyncShardMarkerTrack<rgw_obj_key, rgw_obj_key> { | |
2080 | RGWDataSyncEnv *sync_env; | |
2081 | ||
2082 | string marker_oid; | |
2083 | rgw_bucket_shard_full_sync_marker sync_marker; | |
2084 | ||
2085 | public: | |
2086 | RGWBucketFullSyncShardMarkerTrack(RGWDataSyncEnv *_sync_env, | |
2087 | const string& _marker_oid, | |
2088 | const rgw_bucket_shard_full_sync_marker& _marker) : RGWSyncShardMarkerTrack(BUCKET_SYNC_UPDATE_MARKER_WINDOW), | |
2089 | sync_env(_sync_env), | |
2090 | marker_oid(_marker_oid), | |
2091 | sync_marker(_marker) {} | |
2092 | ||
2093 | RGWCoroutine *store_marker(const rgw_obj_key& new_marker, uint64_t index_pos, const real_time& timestamp) override { | |
2094 | sync_marker.position = new_marker; | |
2095 | sync_marker.count = index_pos; | |
2096 | ||
2097 | map<string, bufferlist> attrs; | |
2098 | sync_marker.encode_attr(attrs); | |
2099 | ||
2100 | RGWRados *store = sync_env->store; | |
2101 | ||
2102 | ldout(sync_env->cct, 20) << __func__ << "(): updating marker marker_oid=" << marker_oid << " marker=" << new_marker << dendl; | |
2103 | return new RGWSimpleRadosWriteAttrsCR(sync_env->async_rados, store, | |
2104 | rgw_raw_obj(store->get_zone_params().log_pool, marker_oid), | |
2105 | attrs); | |
2106 | } | |
2107 | }; | |
2108 | ||
2109 | class RGWBucketIncSyncShardMarkerTrack : public RGWSyncShardMarkerTrack<string, rgw_obj_key> { | |
2110 | RGWDataSyncEnv *sync_env; | |
2111 | ||
2112 | string marker_oid; | |
2113 | rgw_bucket_shard_inc_sync_marker sync_marker; | |
2114 | ||
2115 | map<rgw_obj_key, string> key_to_marker; | |
2116 | map<string, rgw_obj_key> marker_to_key; | |
2117 | ||
2118 | void handle_finish(const string& marker) override { | |
2119 | map<string, rgw_obj_key>::iterator iter = marker_to_key.find(marker); | |
2120 | if (iter == marker_to_key.end()) { | |
2121 | return; | |
2122 | } | |
2123 | key_to_marker.erase(iter->second); | |
2124 | reset_need_retry(iter->second); | |
2125 | marker_to_key.erase(iter); | |
2126 | } | |
2127 | ||
2128 | public: | |
2129 | RGWBucketIncSyncShardMarkerTrack(RGWDataSyncEnv *_sync_env, | |
2130 | const string& _marker_oid, | |
2131 | const rgw_bucket_shard_inc_sync_marker& _marker) : RGWSyncShardMarkerTrack(BUCKET_SYNC_UPDATE_MARKER_WINDOW), | |
2132 | sync_env(_sync_env), | |
2133 | marker_oid(_marker_oid), | |
2134 | sync_marker(_marker) {} | |
2135 | ||
2136 | RGWCoroutine *store_marker(const string& new_marker, uint64_t index_pos, const real_time& timestamp) override { | |
2137 | sync_marker.position = new_marker; | |
2138 | ||
2139 | map<string, bufferlist> attrs; | |
2140 | sync_marker.encode_attr(attrs); | |
2141 | ||
2142 | RGWRados *store = sync_env->store; | |
2143 | ||
2144 | ldout(sync_env->cct, 20) << __func__ << "(): updating marker marker_oid=" << marker_oid << " marker=" << new_marker << dendl; | |
2145 | return new RGWSimpleRadosWriteAttrsCR(sync_env->async_rados, | |
2146 | store, | |
2147 | rgw_raw_obj(store->get_zone_params().log_pool, marker_oid), | |
2148 | attrs); | |
2149 | } | |
2150 | ||
2151 | /* | |
2152 | * create index from key -> <op, marker>, and from marker -> key | |
2153 | * this is useful so that we can insure that we only have one | |
2154 | * entry for any key that is used. This is needed when doing | |
2155 | * incremenatl sync of data, and we don't want to run multiple | |
2156 | * concurrent sync operations for the same bucket shard | |
2157 | * Also, we should make sure that we don't run concurrent operations on the same key with | |
2158 | * different ops. | |
2159 | */ | |
2160 | bool index_key_to_marker(const rgw_obj_key& key, const string& marker) { | |
2161 | if (key_to_marker.find(key) != key_to_marker.end()) { | |
2162 | set_need_retry(key); | |
2163 | return false; | |
2164 | } | |
2165 | key_to_marker[key] = marker; | |
2166 | marker_to_key[marker] = key; | |
2167 | return true; | |
2168 | } | |
2169 | ||
2170 | bool can_do_op(const rgw_obj_key& key) { | |
2171 | return (key_to_marker.find(key) == key_to_marker.end()); | |
2172 | } | |
2173 | }; | |
2174 | ||
2175 | template <class T, class K> | |
2176 | class RGWBucketSyncSingleEntryCR : public RGWCoroutine { | |
2177 | RGWDataSyncEnv *sync_env; | |
2178 | ||
2179 | RGWBucketInfo *bucket_info; | |
2180 | const rgw_bucket_shard& bs; | |
2181 | ||
2182 | rgw_obj_key key; | |
2183 | bool versioned; | |
2184 | uint64_t versioned_epoch; | |
2185 | rgw_bucket_entry_owner owner; | |
2186 | real_time timestamp; | |
2187 | RGWModifyOp op; | |
2188 | RGWPendingState op_state; | |
2189 | ||
2190 | T entry_marker; | |
2191 | RGWSyncShardMarkerTrack<T, K> *marker_tracker; | |
2192 | ||
2193 | int sync_status; | |
2194 | ||
2195 | stringstream error_ss; | |
2196 | ||
2197 | RGWDataSyncDebugLogger logger; | |
2198 | ||
2199 | bool error_injection; | |
2200 | ||
2201 | RGWDataSyncModule *data_sync_module; | |
31f18b77 FG |
2202 | |
2203 | rgw_zone_set zones_trace; | |
7c673cae FG |
2204 | |
2205 | public: | |
2206 | RGWBucketSyncSingleEntryCR(RGWDataSyncEnv *_sync_env, | |
2207 | RGWBucketInfo *_bucket_info, | |
2208 | const rgw_bucket_shard& bs, | |
2209 | const rgw_obj_key& _key, bool _versioned, uint64_t _versioned_epoch, | |
2210 | real_time& _timestamp, | |
2211 | const rgw_bucket_entry_owner& _owner, | |
2212 | RGWModifyOp _op, RGWPendingState _op_state, | |
31f18b77 | 2213 | const T& _entry_marker, RGWSyncShardMarkerTrack<T, K> *_marker_tracker, rgw_zone_set& _zones_trace) : RGWCoroutine(_sync_env->cct), |
7c673cae FG |
2214 | sync_env(_sync_env), |
2215 | bucket_info(_bucket_info), bs(bs), | |
2216 | key(_key), versioned(_versioned), versioned_epoch(_versioned_epoch), | |
2217 | owner(_owner), | |
2218 | timestamp(_timestamp), op(_op), | |
2219 | op_state(_op_state), | |
2220 | entry_marker(_entry_marker), | |
2221 | marker_tracker(_marker_tracker), | |
31f18b77 | 2222 | sync_status(0){ |
7c673cae FG |
2223 | stringstream ss; |
2224 | ss << bucket_shard_str{bs} << "/" << key << "[" << versioned_epoch << "]"; | |
2225 | set_description() << "bucket sync single entry (source_zone=" << sync_env->source_zone << ") b=" << ss.str() << " log_entry=" << entry_marker << " op=" << (int)op << " op_state=" << (int)op_state; | |
2226 | ldout(sync_env->cct, 20) << "bucket sync single entry (source_zone=" << sync_env->source_zone << ") b=" << ss.str() << " log_entry=" << entry_marker << " op=" << (int)op << " op_state=" << (int)op_state << dendl; | |
2227 | set_status("init"); | |
2228 | ||
2229 | logger.init(sync_env, "Object", ss.str()); | |
2230 | ||
2231 | error_injection = (sync_env->cct->_conf->rgw_sync_data_inject_err_probability > 0); | |
2232 | ||
2233 | data_sync_module = sync_env->sync_module->get_data_handler(); | |
31f18b77 FG |
2234 | |
2235 | zones_trace = _zones_trace; | |
2236 | zones_trace.insert(sync_env->store->get_zone().id); | |
7c673cae FG |
2237 | } |
2238 | ||
2239 | int operate() override { | |
2240 | reenter(this) { | |
2241 | /* skip entries that are not complete */ | |
2242 | if (op_state != CLS_RGW_STATE_COMPLETE) { | |
2243 | goto done; | |
2244 | } | |
2245 | do { | |
2246 | yield { | |
2247 | marker_tracker->reset_need_retry(key); | |
2248 | if (key.name.empty()) { | |
2249 | /* shouldn't happen */ | |
2250 | set_status("skipping empty entry"); | |
2251 | ldout(sync_env->cct, 0) << "ERROR: " << __func__ << "(): entry with empty obj name, skipping" << dendl; | |
2252 | goto done; | |
2253 | } | |
2254 | if (error_injection && | |
2255 | rand() % 10000 < cct->_conf->rgw_sync_data_inject_err_probability * 10000.0) { | |
2256 | ldout(sync_env->cct, 0) << __func__ << ": injecting data sync error on key=" << key.name << dendl; | |
2257 | retcode = -EIO; | |
2258 | } else if (op == CLS_RGW_OP_ADD || | |
2259 | op == CLS_RGW_OP_LINK_OLH) { | |
2260 | if (op == CLS_RGW_OP_ADD && !key.instance.empty() && key.instance != "null") { | |
2261 | set_status("skipping entry"); | |
2262 | ldout(sync_env->cct, 10) << "bucket skipping sync obj: " << sync_env->source_zone << "/" << bucket_info->bucket << "/" << key << "[" << versioned_epoch << "]: versioned object will be synced on link_olh" << dendl; | |
2263 | goto done; | |
2264 | ||
2265 | } | |
2266 | set_status("syncing obj"); | |
2267 | ldout(sync_env->cct, 5) << "bucket sync: sync obj: " << sync_env->source_zone << "/" << bucket_info->bucket << "/" << key << "[" << versioned_epoch << "]" << dendl; | |
2268 | logger.log("fetch"); | |
31f18b77 | 2269 | call(data_sync_module->sync_object(sync_env, *bucket_info, key, versioned_epoch, &zones_trace)); |
7c673cae FG |
2270 | } else if (op == CLS_RGW_OP_DEL || op == CLS_RGW_OP_UNLINK_INSTANCE) { |
2271 | set_status("removing obj"); | |
2272 | if (op == CLS_RGW_OP_UNLINK_INSTANCE) { | |
2273 | versioned = true; | |
2274 | } | |
2275 | logger.log("remove"); | |
31f18b77 | 2276 | call(data_sync_module->remove_object(sync_env, *bucket_info, key, timestamp, versioned, versioned_epoch, &zones_trace)); |
7c673cae FG |
2277 | } else if (op == CLS_RGW_OP_LINK_OLH_DM) { |
2278 | logger.log("creating delete marker"); | |
2279 | set_status("creating delete marker"); | |
2280 | ldout(sync_env->cct, 10) << "creating delete marker: obj: " << sync_env->source_zone << "/" << bucket_info->bucket << "/" << key << "[" << versioned_epoch << "]" << dendl; | |
31f18b77 | 2281 | call(data_sync_module->create_delete_marker(sync_env, *bucket_info, key, timestamp, owner, versioned, versioned_epoch, &zones_trace)); |
7c673cae FG |
2282 | } |
2283 | } | |
2284 | } while (marker_tracker->need_retry(key)); | |
2285 | { | |
2286 | stringstream ss; | |
2287 | if (retcode >= 0) { | |
2288 | ss << "done"; | |
2289 | } else { | |
2290 | ss << "done, retcode=" << retcode; | |
2291 | } | |
2292 | logger.log(ss.str()); | |
2293 | } | |
2294 | ||
2295 | if (retcode < 0 && retcode != -ENOENT) { | |
2296 | set_status() << "failed to sync obj; retcode=" << retcode; | |
2297 | ldout(sync_env->cct, 0) << "ERROR: failed to sync object: " | |
2298 | << bucket_shard_str{bs} << "/" << key.name << dendl; | |
2299 | error_ss << bucket_shard_str{bs} << "/" << key.name; | |
2300 | sync_status = retcode; | |
2301 | } | |
2302 | if (!error_ss.str().empty()) { | |
2303 | yield call(sync_env->error_logger->log_error_cr(sync_env->conn->get_remote_id(), "data", error_ss.str(), -retcode, "failed to sync object")); | |
2304 | } | |
2305 | done: | |
2306 | if (sync_status == 0) { | |
2307 | /* update marker */ | |
2308 | set_status() << "calling marker_tracker->finish(" << entry_marker << ")"; | |
2309 | yield call(marker_tracker->finish(entry_marker)); | |
2310 | sync_status = retcode; | |
2311 | } | |
2312 | if (sync_status < 0) { | |
2313 | return set_cr_error(sync_status); | |
2314 | } | |
2315 | return set_cr_done(); | |
2316 | } | |
2317 | return 0; | |
2318 | } | |
2319 | }; | |
2320 | ||
2321 | #define BUCKET_SYNC_SPAWN_WINDOW 20 | |
2322 | ||
2323 | class RGWBucketShardFullSyncCR : public RGWCoroutine { | |
2324 | RGWDataSyncEnv *sync_env; | |
2325 | const rgw_bucket_shard& bs; | |
2326 | RGWBucketInfo *bucket_info; | |
2327 | boost::intrusive_ptr<RGWContinuousLeaseCR> lease_cr; | |
2328 | bucket_list_result list_result; | |
2329 | list<bucket_list_entry>::iterator entries_iter; | |
2330 | rgw_bucket_shard_full_sync_marker& full_marker; | |
2331 | RGWBucketFullSyncShardMarkerTrack marker_tracker; | |
2332 | rgw_obj_key list_marker; | |
2333 | bucket_list_entry *entry{nullptr}; | |
2334 | RGWModifyOp op{CLS_RGW_OP_ADD}; | |
2335 | ||
2336 | int total_entries{0}; | |
2337 | ||
2338 | int sync_status{0}; | |
2339 | ||
2340 | const string& status_oid; | |
2341 | ||
2342 | RGWDataSyncDebugLogger logger; | |
31f18b77 | 2343 | rgw_zone_set zones_trace; |
7c673cae FG |
2344 | public: |
2345 | RGWBucketShardFullSyncCR(RGWDataSyncEnv *_sync_env, const rgw_bucket_shard& bs, | |
2346 | RGWBucketInfo *_bucket_info, | |
2347 | const std::string& status_oid, | |
2348 | RGWContinuousLeaseCR *lease_cr, | |
2349 | rgw_bucket_shard_full_sync_marker& _full_marker) | |
2350 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), bs(bs), | |
2351 | bucket_info(_bucket_info), lease_cr(lease_cr), full_marker(_full_marker), | |
2352 | marker_tracker(sync_env, status_oid, full_marker), | |
2353 | status_oid(status_oid) { | |
2354 | logger.init(sync_env, "BucketFull", bs.get_key()); | |
31f18b77 | 2355 | zones_trace.insert(sync_env->source_zone); |
7c673cae FG |
2356 | } |
2357 | ||
2358 | int operate() override; | |
2359 | }; | |
2360 | ||
2361 | int RGWBucketShardFullSyncCR::operate() | |
2362 | { | |
2363 | int ret; | |
2364 | reenter(this) { | |
2365 | list_marker = full_marker.position; | |
2366 | ||
2367 | total_entries = full_marker.count; | |
2368 | do { | |
2369 | if (!lease_cr->is_locked()) { | |
2370 | drain_all(); | |
2371 | return set_cr_error(-ECANCELED); | |
2372 | } | |
2373 | set_status("listing remote bucket"); | |
2374 | ldout(sync_env->cct, 20) << __func__ << "(): listing bucket for full sync" << dendl; | |
2375 | yield call(new RGWListBucketShardCR(sync_env, bs, list_marker, | |
2376 | &list_result)); | |
2377 | if (retcode < 0 && retcode != -ENOENT) { | |
2378 | set_status("failed bucket listing, going down"); | |
2379 | drain_all(); | |
2380 | return set_cr_error(retcode); | |
2381 | } | |
2382 | entries_iter = list_result.entries.begin(); | |
2383 | for (; entries_iter != list_result.entries.end(); ++entries_iter) { | |
2384 | if (!lease_cr->is_locked()) { | |
2385 | drain_all(); | |
2386 | return set_cr_error(-ECANCELED); | |
2387 | } | |
2388 | ldout(sync_env->cct, 20) << "[full sync] syncing object: " | |
2389 | << bucket_shard_str{bs} << "/" << entries_iter->key << dendl; | |
2390 | entry = &(*entries_iter); | |
2391 | total_entries++; | |
2392 | list_marker = entries_iter->key; | |
2393 | if (!marker_tracker.start(entry->key, total_entries, real_time())) { | |
2394 | ldout(sync_env->cct, 0) << "ERROR: cannot start syncing " << entry->key << ". Duplicate entry?" << dendl; | |
2395 | } else { | |
2396 | op = (entry->key.instance.empty() || entry->key.instance == "null" ? CLS_RGW_OP_ADD : CLS_RGW_OP_LINK_OLH); | |
7c673cae FG |
2397 | using SyncCR = RGWBucketSyncSingleEntryCR<rgw_obj_key, rgw_obj_key>; |
2398 | yield spawn(new SyncCR(sync_env, bucket_info, bs, entry->key, | |
2399 | false, /* versioned, only matters for object removal */ | |
2400 | entry->versioned_epoch, entry->mtime, | |
2401 | entry->owner, op, CLS_RGW_STATE_COMPLETE, | |
31f18b77 | 2402 | entry->key, &marker_tracker, zones_trace), |
7c673cae FG |
2403 | false); |
2404 | } | |
2405 | while (num_spawned() > BUCKET_SYNC_SPAWN_WINDOW) { | |
2406 | yield wait_for_child(); | |
2407 | bool again = true; | |
2408 | while (again) { | |
2409 | again = collect(&ret, nullptr); | |
2410 | if (ret < 0) { | |
2411 | ldout(sync_env->cct, 0) << "ERROR: a sync operation returned error" << dendl; | |
2412 | sync_status = ret; | |
2413 | /* we have reported this error */ | |
2414 | } | |
2415 | } | |
2416 | } | |
2417 | } | |
2418 | } while (list_result.is_truncated && sync_status == 0); | |
2419 | set_status("done iterating over all objects"); | |
2420 | /* wait for all operations to complete */ | |
2421 | while (num_spawned()) { | |
2422 | yield wait_for_child(); | |
2423 | bool again = true; | |
2424 | while (again) { | |
2425 | again = collect(&ret, nullptr); | |
2426 | if (ret < 0) { | |
2427 | ldout(sync_env->cct, 0) << "ERROR: a sync operation returned error" << dendl; | |
2428 | sync_status = ret; | |
2429 | /* we have reported this error */ | |
2430 | } | |
2431 | } | |
2432 | } | |
2433 | if (!lease_cr->is_locked()) { | |
2434 | return set_cr_error(-ECANCELED); | |
2435 | } | |
2436 | /* update sync state to incremental */ | |
2437 | if (sync_status == 0) { | |
2438 | yield { | |
2439 | rgw_bucket_shard_sync_info sync_status; | |
2440 | sync_status.state = rgw_bucket_shard_sync_info::StateIncrementalSync; | |
2441 | map<string, bufferlist> attrs; | |
2442 | sync_status.encode_state_attr(attrs); | |
2443 | RGWRados *store = sync_env->store; | |
2444 | call(new RGWSimpleRadosWriteAttrsCR(sync_env->async_rados, store, | |
2445 | rgw_raw_obj(store->get_zone_params().log_pool, status_oid), | |
2446 | attrs)); | |
2447 | } | |
2448 | } else { | |
2449 | ldout(sync_env->cct, 0) << "ERROR: failure in sync, backing out (sync_status=" << sync_status<< ")" << dendl; | |
2450 | } | |
2451 | if (retcode < 0 && sync_status == 0) { /* actually tried to set incremental state and failed */ | |
2452 | ldout(sync_env->cct, 0) << "ERROR: failed to set sync state on bucket " | |
2453 | << bucket_shard_str{bs} << " retcode=" << retcode << dendl; | |
2454 | return set_cr_error(retcode); | |
2455 | } | |
2456 | if (sync_status < 0) { | |
2457 | return set_cr_error(sync_status); | |
2458 | } | |
2459 | return set_cr_done(); | |
2460 | } | |
2461 | return 0; | |
2462 | } | |
2463 | ||
2464 | class RGWBucketShardIncrementalSyncCR : public RGWCoroutine { | |
2465 | RGWDataSyncEnv *sync_env; | |
2466 | const rgw_bucket_shard& bs; | |
2467 | RGWBucketInfo *bucket_info; | |
2468 | boost::intrusive_ptr<RGWContinuousLeaseCR> lease_cr; | |
2469 | list<rgw_bi_log_entry> list_result; | |
2470 | list<rgw_bi_log_entry>::iterator entries_iter; | |
2471 | map<pair<string, string>, pair<real_time, RGWModifyOp> > squash_map; | |
2472 | rgw_bucket_shard_inc_sync_marker& inc_marker; | |
2473 | rgw_obj_key key; | |
2474 | rgw_bi_log_entry *entry{nullptr}; | |
2475 | RGWBucketIncSyncShardMarkerTrack marker_tracker; | |
2476 | bool updated_status{false}; | |
2477 | const string& status_oid; | |
31f18b77 | 2478 | const string& zone_id; |
7c673cae FG |
2479 | |
2480 | string cur_id; | |
2481 | ||
2482 | RGWDataSyncDebugLogger logger; | |
2483 | ||
2484 | int sync_status{0}; | |
2485 | ||
2486 | public: | |
2487 | RGWBucketShardIncrementalSyncCR(RGWDataSyncEnv *_sync_env, | |
2488 | const rgw_bucket_shard& bs, | |
2489 | RGWBucketInfo *_bucket_info, | |
2490 | const std::string& status_oid, | |
2491 | RGWContinuousLeaseCR *lease_cr, | |
2492 | rgw_bucket_shard_inc_sync_marker& _inc_marker) | |
2493 | : RGWCoroutine(_sync_env->cct), sync_env(_sync_env), bs(bs), | |
2494 | bucket_info(_bucket_info), lease_cr(lease_cr), inc_marker(_inc_marker), | |
31f18b77 | 2495 | marker_tracker(sync_env, status_oid, inc_marker), status_oid(status_oid) , zone_id(_sync_env->store->get_zone().id){ |
7c673cae FG |
2496 | set_description() << "bucket shard incremental sync bucket=" |
2497 | << bucket_shard_str{bs}; | |
2498 | set_status("init"); | |
2499 | logger.init(sync_env, "BucketInc", bs.get_key()); | |
2500 | } | |
2501 | ||
2502 | int operate() override; | |
2503 | }; | |
2504 | ||
2505 | int RGWBucketShardIncrementalSyncCR::operate() | |
2506 | { | |
2507 | int ret; | |
2508 | reenter(this) { | |
2509 | do { | |
2510 | if (!lease_cr->is_locked()) { | |
2511 | drain_all(); | |
2512 | return set_cr_error(-ECANCELED); | |
2513 | } | |
2514 | ldout(sync_env->cct, 20) << __func__ << "(): listing bilog for incremental sync" << dendl; | |
2515 | set_status() << "listing bilog; position=" << inc_marker.position; | |
2516 | yield call(new RGWListBucketIndexLogCR(sync_env, bs, inc_marker.position, | |
2517 | &list_result)); | |
2518 | if (retcode < 0 && retcode != -ENOENT) { | |
2519 | /* wait for all operations to complete */ | |
2520 | drain_all(); | |
2521 | return set_cr_error(retcode); | |
2522 | } | |
2523 | squash_map.clear(); | |
2524 | for (auto& e : list_result) { | |
2525 | if (e.state != CLS_RGW_STATE_COMPLETE) { | |
2526 | continue; | |
2527 | } | |
31f18b77 FG |
2528 | if (e.zones_trace.find(zone_id) != e.zones_trace.end()) { |
2529 | continue; | |
2530 | } | |
7c673cae FG |
2531 | auto& squash_entry = squash_map[make_pair(e.object, e.instance)]; |
2532 | if (squash_entry.first <= e.timestamp) { | |
2533 | squash_entry = make_pair<>(e.timestamp, e.op); | |
2534 | } | |
2535 | } | |
2536 | entries_iter = list_result.begin(); | |
2537 | for (; entries_iter != list_result.end(); ++entries_iter) { | |
2538 | if (!lease_cr->is_locked()) { | |
2539 | drain_all(); | |
2540 | return set_cr_error(-ECANCELED); | |
2541 | } | |
2542 | entry = &(*entries_iter); | |
2543 | { | |
2544 | ssize_t p = entry->id.find('#'); /* entries might have explicit shard info in them, e.g., 6#00000000004.94.3 */ | |
2545 | if (p < 0) { | |
2546 | cur_id = entry->id; | |
2547 | } else { | |
2548 | cur_id = entry->id.substr(p + 1); | |
2549 | } | |
2550 | } | |
2551 | inc_marker.position = cur_id; | |
2552 | ||
2553 | if (!key.set(rgw_obj_index_key{entry->object, entry->instance})) { | |
2554 | set_status() << "parse_raw_oid() on " << entry->object << " returned false, skipping entry"; | |
2555 | ldout(sync_env->cct, 20) << "parse_raw_oid() on " << entry->object << " returned false, skipping entry" << dendl; | |
2556 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2557 | continue; | |
2558 | } | |
2559 | ||
2560 | ldout(sync_env->cct, 20) << "parsed entry: id=" << cur_id << " iter->object=" << entry->object << " iter->instance=" << entry->instance << " name=" << key.name << " instance=" << key.instance << " ns=" << key.ns << dendl; | |
2561 | ||
2562 | if (!key.ns.empty()) { | |
2563 | set_status() << "skipping entry in namespace: " << entry->object; | |
2564 | ldout(sync_env->cct, 20) << "skipping entry in namespace: " << entry->object << dendl; | |
2565 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2566 | continue; | |
2567 | } | |
2568 | ||
2569 | set_status() << "got entry.id=" << cur_id << " key=" << key << " op=" << (int)entry->op; | |
2570 | if (entry->op == CLS_RGW_OP_CANCEL) { | |
2571 | set_status() << "canceled operation, skipping"; | |
2572 | ldout(sync_env->cct, 20) << "[inc sync] skipping object: " | |
2573 | << bucket_shard_str{bs} << "/" << key << ": canceled operation" << dendl; | |
2574 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2575 | continue; | |
2576 | } | |
2577 | if (entry->state != CLS_RGW_STATE_COMPLETE) { | |
2578 | set_status() << "non-complete operation, skipping"; | |
2579 | ldout(sync_env->cct, 20) << "[inc sync] skipping object: " | |
2580 | << bucket_shard_str{bs} << "/" << key << ": non-complete operation" << dendl; | |
2581 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2582 | continue; | |
2583 | } | |
31f18b77 FG |
2584 | if (entry->zones_trace.find(zone_id) != entry->zones_trace.end()) { |
2585 | set_status() << "redundant operation, skipping"; | |
2586 | ldout(sync_env->cct, 20) << "[inc sync] skipping object: " | |
2587 | <<bucket_shard_str{bs} <<"/"<<key<<": redundant operation" << dendl; | |
2588 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2589 | continue; | |
2590 | } | |
7c673cae FG |
2591 | if (make_pair<>(entry->timestamp, entry->op) != squash_map[make_pair(entry->object, entry->instance)]) { |
2592 | set_status() << "squashed operation, skipping"; | |
2593 | ldout(sync_env->cct, 20) << "[inc sync] skipping object: " | |
2594 | << bucket_shard_str{bs} << "/" << key << ": squashed operation" << dendl; | |
2595 | /* not updating high marker though */ | |
2596 | continue; | |
2597 | } | |
2598 | ldout(sync_env->cct, 20) << "[inc sync] syncing object: " | |
2599 | << bucket_shard_str{bs} << "/" << key << dendl; | |
2600 | updated_status = false; | |
2601 | while (!marker_tracker.can_do_op(key)) { | |
2602 | if (!updated_status) { | |
2603 | set_status() << "can't do op, conflicting inflight operation"; | |
2604 | updated_status = true; | |
2605 | } | |
2606 | ldout(sync_env->cct, 5) << *this << ": [inc sync] can't do op on key=" << key << " need to wait for conflicting operation to complete" << dendl; | |
2607 | yield wait_for_child(); | |
2608 | bool again = true; | |
2609 | while (again) { | |
2610 | again = collect(&ret, nullptr); | |
2611 | if (ret < 0) { | |
2612 | ldout(sync_env->cct, 0) << "ERROR: a child operation returned error (ret=" << ret << ")" << dendl; | |
2613 | sync_status = ret; | |
2614 | /* we have reported this error */ | |
2615 | } | |
2616 | } | |
2617 | } | |
2618 | if (!marker_tracker.index_key_to_marker(key, cur_id)) { | |
2619 | set_status() << "can't do op, sync already in progress for object"; | |
2620 | ldout(sync_env->cct, 20) << __func__ << ": skipping sync of entry: " << cur_id << ":" << key << " sync already in progress for object" << dendl; | |
2621 | marker_tracker.try_update_high_marker(cur_id, 0, entry->timestamp); | |
2622 | continue; | |
2623 | } | |
2624 | // yield { | |
2625 | set_status() << "start object sync"; | |
2626 | if (!marker_tracker.start(cur_id, 0, entry->timestamp)) { | |
2627 | ldout(sync_env->cct, 0) << "ERROR: cannot start syncing " << cur_id << ". Duplicate entry?" << dendl; | |
2628 | } else { | |
2629 | uint64_t versioned_epoch = 0; | |
2630 | rgw_bucket_entry_owner owner(entry->owner, entry->owner_display_name); | |
2631 | if (entry->ver.pool < 0) { | |
2632 | versioned_epoch = entry->ver.epoch; | |
2633 | } | |
2634 | ldout(sync_env->cct, 20) << __func__ << "(): entry->timestamp=" << entry->timestamp << dendl; | |
2635 | using SyncCR = RGWBucketSyncSingleEntryCR<string, rgw_obj_key>; | |
2636 | spawn(new SyncCR(sync_env, bucket_info, bs, key, | |
2637 | entry->is_versioned(), versioned_epoch, | |
2638 | entry->timestamp, owner, entry->op, entry->state, | |
31f18b77 | 2639 | cur_id, &marker_tracker, entry->zones_trace), |
7c673cae FG |
2640 | false); |
2641 | } | |
2642 | // } | |
2643 | while (num_spawned() > BUCKET_SYNC_SPAWN_WINDOW) { | |
2644 | set_status() << "num_spawned() > spawn_window"; | |
2645 | yield wait_for_child(); | |
2646 | bool again = true; | |
2647 | while (again) { | |
2648 | again = collect(&ret, nullptr); | |
2649 | if (ret < 0) { | |
2650 | ldout(sync_env->cct, 0) << "ERROR: a sync operation returned error" << dendl; | |
2651 | sync_status = ret; | |
2652 | /* we have reported this error */ | |
2653 | } | |
2654 | /* not waiting for child here */ | |
2655 | } | |
2656 | } | |
2657 | } | |
2658 | } while (!list_result.empty() && sync_status == 0); | |
2659 | ||
2660 | while (num_spawned()) { | |
2661 | yield wait_for_child(); | |
2662 | bool again = true; | |
2663 | while (again) { | |
2664 | again = collect(&ret, nullptr); | |
2665 | if (ret < 0) { | |
2666 | ldout(sync_env->cct, 0) << "ERROR: a sync operation returned error" << dendl; | |
2667 | sync_status = ret; | |
2668 | /* we have reported this error */ | |
2669 | } | |
2670 | /* not waiting for child here */ | |
2671 | } | |
2672 | } | |
2673 | ||
2674 | yield call(marker_tracker.flush()); | |
2675 | if (retcode < 0) { | |
2676 | ldout(sync_env->cct, 0) << "ERROR: marker_tracker.flush() returned retcode=" << retcode << dendl; | |
2677 | return set_cr_error(retcode); | |
2678 | } | |
2679 | if (sync_status < 0) { | |
2680 | ldout(sync_env->cct, 0) << "ERROR: failure in sync, backing out (sync_status=" << sync_status<< ")" << dendl; | |
2681 | } | |
2682 | ||
2683 | /* wait for all operations to complete */ | |
2684 | drain_all(); | |
2685 | ||
2686 | if (sync_status < 0) { | |
2687 | return set_cr_error(sync_status); | |
2688 | } | |
2689 | ||
2690 | return set_cr_done(); | |
2691 | } | |
2692 | return 0; | |
2693 | } | |
2694 | ||
2695 | int RGWRunBucketSyncCoroutine::operate() | |
2696 | { | |
2697 | reenter(this) { | |
2698 | yield { | |
2699 | set_status("acquiring sync lock"); | |
2700 | auto store = sync_env->store; | |
31f18b77 FG |
2701 | lease_cr.reset(new RGWContinuousLeaseCR(sync_env->async_rados, store, |
2702 | rgw_raw_obj(store->get_zone_params().log_pool, status_oid), | |
2703 | "sync_lock", | |
2704 | cct->_conf->rgw_sync_lease_period, | |
2705 | this)); | |
2706 | lease_stack.reset(spawn(lease_cr.get(), false)); | |
7c673cae FG |
2707 | } |
2708 | while (!lease_cr->is_locked()) { | |
2709 | if (lease_cr->is_done()) { | |
2710 | ldout(cct, 5) << "lease cr failed, done early" << dendl; | |
2711 | set_status("lease lock failed, early abort"); | |
2712 | return set_cr_error(lease_cr->get_ret_status()); | |
2713 | } | |
2714 | set_sleeping(true); | |
2715 | yield; | |
2716 | } | |
2717 | ||
2718 | yield call(new RGWReadBucketSyncStatusCoroutine(sync_env, bs, &sync_status)); | |
2719 | if (retcode < 0 && retcode != -ENOENT) { | |
2720 | ldout(sync_env->cct, 0) << "ERROR: failed to read sync status for bucket=" | |
2721 | << bucket_shard_str{bs} << dendl; | |
2722 | lease_cr->go_down(); | |
2723 | drain_all(); | |
2724 | return set_cr_error(retcode); | |
2725 | } | |
2726 | ||
2727 | ldout(sync_env->cct, 20) << __func__ << "(): sync status for bucket " | |
2728 | << bucket_shard_str{bs} << ": " << sync_status.state << dendl; | |
2729 | ||
2730 | yield call(new RGWGetBucketInstanceInfoCR(sync_env->async_rados, sync_env->store, bs.bucket, &bucket_info)); | |
2731 | if (retcode == -ENOENT) { | |
2732 | /* bucket instance info has not been synced in yet, fetch it now */ | |
2733 | yield { | |
2734 | ldout(sync_env->cct, 10) << "no local info for bucket " | |
2735 | << bucket_str{bs.bucket} << ": fetching metadata" << dendl; | |
2736 | string raw_key = string("bucket.instance:") + bs.bucket.get_key(); | |
2737 | ||
2738 | meta_sync_env.init(cct, sync_env->store, sync_env->store->rest_master_conn, sync_env->async_rados, sync_env->http_manager, sync_env->error_logger); | |
2739 | ||
2740 | call(new RGWMetaSyncSingleEntryCR(&meta_sync_env, raw_key, | |
2741 | string() /* no marker */, | |
2742 | MDLOG_STATUS_COMPLETE, | |
2743 | NULL /* no marker tracker */)); | |
2744 | } | |
2745 | if (retcode < 0) { | |
2746 | ldout(sync_env->cct, 0) << "ERROR: failed to fetch bucket instance info for " << bucket_str{bs.bucket} << dendl; | |
2747 | lease_cr->go_down(); | |
2748 | drain_all(); | |
2749 | return set_cr_error(retcode); | |
2750 | } | |
2751 | ||
2752 | yield call(new RGWGetBucketInstanceInfoCR(sync_env->async_rados, sync_env->store, bs.bucket, &bucket_info)); | |
2753 | } | |
2754 | if (retcode < 0) { | |
2755 | ldout(sync_env->cct, 0) << "ERROR: failed to retrieve bucket info for bucket=" << bucket_str{bs.bucket} << dendl; | |
2756 | lease_cr->go_down(); | |
2757 | drain_all(); | |
2758 | return set_cr_error(retcode); | |
2759 | } | |
2760 | ||
2761 | if (sync_status.state == rgw_bucket_shard_sync_info::StateInit) { | |
2762 | yield call(new RGWInitBucketShardSyncStatusCoroutine(sync_env, bs, sync_status)); | |
2763 | if (retcode < 0) { | |
2764 | ldout(sync_env->cct, 0) << "ERROR: init sync on " << bucket_shard_str{bs} | |
2765 | << " failed, retcode=" << retcode << dendl; | |
2766 | lease_cr->go_down(); | |
2767 | drain_all(); | |
2768 | return set_cr_error(retcode); | |
2769 | } | |
2770 | } | |
2771 | ||
2772 | if (sync_status.state == rgw_bucket_shard_sync_info::StateFullSync) { | |
2773 | yield call(new RGWBucketShardFullSyncCR(sync_env, bs, &bucket_info, | |
2774 | status_oid, lease_cr.get(), | |
2775 | sync_status.full_marker)); | |
2776 | if (retcode < 0) { | |
2777 | ldout(sync_env->cct, 5) << "full sync on " << bucket_shard_str{bs} | |
2778 | << " failed, retcode=" << retcode << dendl; | |
2779 | lease_cr->go_down(); | |
2780 | drain_all(); | |
2781 | return set_cr_error(retcode); | |
2782 | } | |
2783 | sync_status.state = rgw_bucket_shard_sync_info::StateIncrementalSync; | |
2784 | } | |
2785 | ||
2786 | if (sync_status.state == rgw_bucket_shard_sync_info::StateIncrementalSync) { | |
2787 | yield call(new RGWBucketShardIncrementalSyncCR(sync_env, bs, &bucket_info, | |
2788 | status_oid, lease_cr.get(), | |
2789 | sync_status.inc_marker)); | |
2790 | if (retcode < 0) { | |
2791 | ldout(sync_env->cct, 5) << "incremental sync on " << bucket_shard_str{bs} | |
2792 | << " failed, retcode=" << retcode << dendl; | |
2793 | lease_cr->go_down(); | |
2794 | drain_all(); | |
2795 | return set_cr_error(retcode); | |
2796 | } | |
2797 | } | |
2798 | ||
2799 | lease_cr->go_down(); | |
2800 | drain_all(); | |
2801 | return set_cr_done(); | |
2802 | } | |
2803 | ||
2804 | return 0; | |
2805 | } | |
2806 | ||
2807 | RGWCoroutine *RGWRemoteBucketLog::run_sync_cr() | |
2808 | { | |
2809 | return new RGWRunBucketSyncCoroutine(&sync_env, bs); | |
2810 | } | |
2811 | ||
2812 | int RGWBucketSyncStatusManager::init() | |
2813 | { | |
2814 | conn = store->get_zone_conn_by_id(source_zone); | |
2815 | if (!conn) { | |
2816 | ldout(store->ctx(), 0) << "connection object to zone " << source_zone << " does not exist" << dendl; | |
2817 | return -EINVAL; | |
2818 | } | |
2819 | ||
2820 | int ret = http_manager.set_threaded(); | |
2821 | if (ret < 0) { | |
2822 | ldout(store->ctx(), 0) << "failed in http_manager.set_threaded() ret=" << ret << dendl; | |
2823 | return ret; | |
2824 | } | |
2825 | ||
2826 | ||
2827 | const string key = bucket.get_key(); | |
2828 | ||
2829 | rgw_http_param_pair pairs[] = { { "key", key.c_str() }, | |
2830 | { NULL, NULL } }; | |
2831 | ||
2832 | string path = string("/admin/metadata/bucket.instance"); | |
2833 | ||
2834 | bucket_instance_meta_info result; | |
2835 | ret = cr_mgr.run(new RGWReadRESTResourceCR<bucket_instance_meta_info>(store->ctx(), conn, &http_manager, path, pairs, &result)); | |
2836 | if (ret < 0) { | |
2837 | ldout(store->ctx(), 0) << "ERROR: failed to fetch bucket metadata info from zone=" << source_zone << " path=" << path << " key=" << key << " ret=" << ret << dendl; | |
2838 | return ret; | |
2839 | } | |
2840 | ||
2841 | RGWBucketInfo& bi = result.data.get_bucket_info(); | |
2842 | num_shards = bi.num_shards; | |
2843 | ||
2844 | error_logger = new RGWSyncErrorLogger(store, RGW_SYNC_ERROR_LOG_SHARD_PREFIX, ERROR_LOGGER_SHARDS); | |
2845 | ||
2846 | sync_module.reset(new RGWDefaultSyncModuleInstance()); | |
2847 | ||
2848 | int effective_num_shards = (num_shards ? num_shards : 1); | |
2849 | ||
2850 | auto async_rados = store->get_async_rados(); | |
2851 | ||
2852 | for (int i = 0; i < effective_num_shards; i++) { | |
2853 | RGWRemoteBucketLog *l = new RGWRemoteBucketLog(store, this, async_rados, &http_manager); | |
2854 | ret = l->init(source_zone, conn, bucket, (num_shards ? i : -1), error_logger, sync_module); | |
2855 | if (ret < 0) { | |
2856 | ldout(store->ctx(), 0) << "ERROR: failed to initialize RGWRemoteBucketLog object" << dendl; | |
2857 | return ret; | |
2858 | } | |
2859 | source_logs[i] = l; | |
2860 | } | |
2861 | ||
2862 | return 0; | |
2863 | } | |
2864 | ||
2865 | int RGWBucketSyncStatusManager::init_sync_status() | |
2866 | { | |
2867 | list<RGWCoroutinesStack *> stacks; | |
2868 | ||
2869 | for (map<int, RGWRemoteBucketLog *>::iterator iter = source_logs.begin(); iter != source_logs.end(); ++iter) { | |
2870 | RGWCoroutinesStack *stack = new RGWCoroutinesStack(store->ctx(), &cr_mgr); | |
2871 | RGWRemoteBucketLog *l = iter->second; | |
2872 | stack->call(l->init_sync_status_cr()); | |
2873 | ||
2874 | stacks.push_back(stack); | |
2875 | } | |
2876 | ||
2877 | return cr_mgr.run(stacks); | |
2878 | } | |
2879 | ||
2880 | int RGWBucketSyncStatusManager::read_sync_status() | |
2881 | { | |
2882 | list<RGWCoroutinesStack *> stacks; | |
2883 | ||
2884 | for (map<int, RGWRemoteBucketLog *>::iterator iter = source_logs.begin(); iter != source_logs.end(); ++iter) { | |
2885 | RGWCoroutinesStack *stack = new RGWCoroutinesStack(store->ctx(), &cr_mgr); | |
2886 | RGWRemoteBucketLog *l = iter->second; | |
2887 | stack->call(l->read_sync_status_cr(&sync_status[iter->first])); | |
2888 | ||
2889 | stacks.push_back(stack); | |
2890 | } | |
2891 | ||
2892 | int ret = cr_mgr.run(stacks); | |
2893 | if (ret < 0) { | |
2894 | ldout(store->ctx(), 0) << "ERROR: failed to read sync status for " | |
2895 | << bucket_str{bucket} << dendl; | |
2896 | return ret; | |
2897 | } | |
2898 | ||
2899 | return 0; | |
2900 | } | |
2901 | ||
2902 | int RGWBucketSyncStatusManager::run() | |
2903 | { | |
2904 | list<RGWCoroutinesStack *> stacks; | |
2905 | ||
2906 | for (map<int, RGWRemoteBucketLog *>::iterator iter = source_logs.begin(); iter != source_logs.end(); ++iter) { | |
2907 | RGWCoroutinesStack *stack = new RGWCoroutinesStack(store->ctx(), &cr_mgr); | |
2908 | RGWRemoteBucketLog *l = iter->second; | |
2909 | stack->call(l->run_sync_cr()); | |
2910 | ||
2911 | stacks.push_back(stack); | |
2912 | } | |
2913 | ||
2914 | int ret = cr_mgr.run(stacks); | |
2915 | if (ret < 0) { | |
2916 | ldout(store->ctx(), 0) << "ERROR: failed to read sync status for " | |
2917 | << bucket_str{bucket} << dendl; | |
2918 | return ret; | |
2919 | } | |
2920 | ||
2921 | return 0; | |
2922 | } | |
2923 | ||
2924 | string RGWBucketSyncStatusManager::status_oid(const string& source_zone, | |
2925 | const rgw_bucket_shard& bs) | |
2926 | { | |
2927 | return bucket_status_oid_prefix + "." + source_zone + ":" + bs.get_key(); | |
2928 | } | |
2929 | ||
2930 | ||
2931 | // TODO: move into rgw_data_sync_trim.cc | |
2932 | #undef dout_prefix | |
2933 | #define dout_prefix (*_dout << "data trim: ") | |
2934 | ||
2935 | namespace { | |
2936 | ||
2937 | /// return the marker that it's safe to trim up to | |
2938 | const std::string& get_stable_marker(const rgw_data_sync_marker& m) | |
2939 | { | |
2940 | return m.state == m.FullSync ? m.next_step_marker : m.marker; | |
2941 | } | |
2942 | ||
2943 | /// comparison operator for take_min_markers() | |
2944 | bool operator<(const rgw_data_sync_marker& lhs, | |
2945 | const rgw_data_sync_marker& rhs) | |
2946 | { | |
2947 | // sort by stable marker | |
2948 | return get_stable_marker(lhs) < get_stable_marker(rhs); | |
2949 | } | |
2950 | ||
2951 | /// populate the container starting with 'dest' with the minimum stable marker | |
2952 | /// of each shard for all of the peers in [first, last) | |
2953 | template <typename IterIn, typename IterOut> | |
2954 | void take_min_markers(IterIn first, IterIn last, IterOut dest) | |
2955 | { | |
2956 | if (first == last) { | |
2957 | return; | |
2958 | } | |
2959 | // initialize markers with the first peer's | |
2960 | auto m = dest; | |
2961 | for (auto &shard : first->sync_markers) { | |
2962 | *m = std::move(shard.second); | |
2963 | ++m; | |
2964 | } | |
2965 | // for remaining peers, replace with smaller markers | |
2966 | for (auto p = first + 1; p != last; ++p) { | |
2967 | m = dest; | |
2968 | for (auto &shard : p->sync_markers) { | |
2969 | if (shard.second < *m) { | |
2970 | *m = std::move(shard.second); | |
2971 | } | |
2972 | ++m; | |
2973 | } | |
2974 | } | |
2975 | } | |
2976 | ||
2977 | } // anonymous namespace | |
2978 | ||
2979 | class DataLogTrimCR : public RGWCoroutine { | |
2980 | RGWRados *store; | |
2981 | RGWHTTPManager *http; | |
2982 | const int num_shards; | |
2983 | const std::string& zone_id; //< my zone id | |
2984 | std::vector<rgw_data_sync_status> peer_status; //< sync status for each peer | |
2985 | std::vector<rgw_data_sync_marker> min_shard_markers; //< min marker per shard | |
2986 | std::vector<std::string>& last_trim; //< last trimmed marker per shard | |
2987 | int ret{0}; | |
2988 | ||
2989 | public: | |
2990 | DataLogTrimCR(RGWRados *store, RGWHTTPManager *http, | |
2991 | int num_shards, std::vector<std::string>& last_trim) | |
2992 | : RGWCoroutine(store->ctx()), store(store), http(http), | |
2993 | num_shards(num_shards), | |
2994 | zone_id(store->get_zone().id), | |
2995 | peer_status(store->zone_conn_map.size()), | |
2996 | min_shard_markers(num_shards), | |
2997 | last_trim(last_trim) | |
2998 | {} | |
2999 | ||
3000 | int operate() override; | |
3001 | }; | |
3002 | ||
3003 | int DataLogTrimCR::operate() | |
3004 | { | |
3005 | reenter(this) { | |
3006 | ldout(cct, 10) << "fetching sync status for zone " << zone_id << dendl; | |
3007 | set_status("fetching sync status"); | |
3008 | yield { | |
3009 | // query data sync status from each sync peer | |
3010 | rgw_http_param_pair params[] = { | |
3011 | { "type", "data" }, | |
3012 | { "status", nullptr }, | |
3013 | { "source-zone", zone_id.c_str() }, | |
3014 | { nullptr, nullptr } | |
3015 | }; | |
3016 | ||
3017 | auto p = peer_status.begin(); | |
3018 | for (auto& c : store->zone_conn_map) { | |
3019 | ldout(cct, 20) << "query sync status from " << c.first << dendl; | |
3020 | using StatusCR = RGWReadRESTResourceCR<rgw_data_sync_status>; | |
3021 | spawn(new StatusCR(cct, c.second, http, "/admin/log/", params, &*p), | |
3022 | false); | |
3023 | ++p; | |
3024 | } | |
3025 | } | |
3026 | ||
3027 | // must get a successful reply from all peers to consider trimming | |
3028 | ret = 0; | |
3029 | while (ret == 0 && num_spawned() > 0) { | |
3030 | yield wait_for_child(); | |
3031 | collect_next(&ret); | |
3032 | } | |
3033 | drain_all(); | |
3034 | ||
3035 | if (ret < 0) { | |
3036 | ldout(cct, 4) << "failed to fetch sync status from all peers" << dendl; | |
3037 | return set_cr_error(ret); | |
3038 | } | |
3039 | ||
3040 | ldout(cct, 10) << "trimming log shards" << dendl; | |
3041 | set_status("trimming log shards"); | |
3042 | yield { | |
3043 | // determine the minimum marker for each shard | |
3044 | take_min_markers(peer_status.begin(), peer_status.end(), | |
3045 | min_shard_markers.begin()); | |
3046 | ||
3047 | for (int i = 0; i < num_shards; i++) { | |
3048 | const auto& m = min_shard_markers[i]; | |
3049 | auto& stable = get_stable_marker(m); | |
3050 | if (stable <= last_trim[i]) { | |
3051 | continue; | |
3052 | } | |
3053 | ldout(cct, 10) << "trimming log shard " << i | |
3054 | << " at marker=" << stable | |
3055 | << " last_trim=" << last_trim[i] << dendl; | |
3056 | using TrimCR = RGWSyncLogTrimCR; | |
3057 | spawn(new TrimCR(store, store->data_log->get_oid(i), | |
3058 | stable, &last_trim[i]), | |
3059 | true); | |
3060 | } | |
3061 | } | |
3062 | return set_cr_done(); | |
3063 | } | |
3064 | return 0; | |
3065 | } | |
3066 | ||
3067 | class DataLogTrimPollCR : public RGWCoroutine { | |
3068 | RGWRados *store; | |
3069 | RGWHTTPManager *http; | |
3070 | const int num_shards; | |
3071 | const utime_t interval; //< polling interval | |
3072 | const std::string lock_oid; //< use first data log shard for lock | |
3073 | const std::string lock_cookie; | |
3074 | std::vector<std::string> last_trim; //< last trimmed marker per shard | |
3075 | ||
3076 | public: | |
3077 | DataLogTrimPollCR(RGWRados *store, RGWHTTPManager *http, | |
3078 | int num_shards, utime_t interval) | |
3079 | : RGWCoroutine(store->ctx()), store(store), http(http), | |
3080 | num_shards(num_shards), interval(interval), | |
3081 | lock_oid(store->data_log->get_oid(0)), | |
3082 | lock_cookie(RGWSimpleRadosLockCR::gen_random_cookie(cct)), | |
3083 | last_trim(num_shards) | |
3084 | {} | |
3085 | ||
3086 | int operate() override; | |
3087 | }; | |
3088 | ||
3089 | int DataLogTrimPollCR::operate() | |
3090 | { | |
3091 | reenter(this) { | |
3092 | for (;;) { | |
3093 | set_status("sleeping"); | |
3094 | wait(interval); | |
3095 | ||
3096 | // request a 'data_trim' lock that covers the entire wait interval to | |
3097 | // prevent other gateways from attempting to trim for the duration | |
3098 | set_status("acquiring trim lock"); | |
3099 | yield call(new RGWSimpleRadosLockCR(store->get_async_rados(), store, | |
3100 | rgw_raw_obj(store->get_zone_params().log_pool, lock_oid), | |
3101 | "data_trim", lock_cookie, | |
3102 | interval.sec())); | |
3103 | if (retcode < 0) { | |
3104 | // if the lock is already held, go back to sleep and try again later | |
3105 | ldout(cct, 4) << "failed to lock " << lock_oid << ", trying again in " | |
3106 | << interval.sec() << "s" << dendl; | |
3107 | continue; | |
3108 | } | |
3109 | ||
3110 | set_status("trimming"); | |
3111 | yield call(new DataLogTrimCR(store, http, num_shards, last_trim)); | |
3112 | ||
3113 | // note that the lock is not released. this is intentional, as it avoids | |
3114 | // duplicating this work in other gateways | |
3115 | } | |
3116 | } | |
3117 | return 0; | |
3118 | } | |
3119 | ||
3120 | RGWCoroutine* create_data_log_trim_cr(RGWRados *store, | |
3121 | RGWHTTPManager *http, | |
3122 | int num_shards, utime_t interval) | |
3123 | { | |
3124 | return new DataLogTrimPollCR(store, http, num_shards, interval); | |
3125 | } |