]>
Commit | Line | Data |
---|---|---|
1 | ========= | |
2 | Migration | |
3 | ========= | |
4 | ||
5 | QEMU has code to load/save the state of the guest that it is running. | |
6 | These are two complementary operations. Saving the state just does | |
7 | that, saves the state for each device that the guest is running. | |
8 | Restoring a guest is just the opposite operation: we need to load the | |
9 | state of each device. | |
10 | ||
11 | For this to work, QEMU has to be launched with the same arguments the | |
12 | two times. I.e. it can only restore the state in one guest that has | |
13 | the same devices that the one it was saved (this last requirement can | |
14 | be relaxed a bit, but for now we can consider that configuration has | |
15 | to be exactly the same). | |
16 | ||
17 | Once that we are able to save/restore a guest, a new functionality is | |
18 | requested: migration. This means that QEMU is able to start in one | |
19 | machine and being "migrated" to another machine. I.e. being moved to | |
20 | another machine. | |
21 | ||
22 | Next was the "live migration" functionality. This is important | |
23 | because some guests run with a lot of state (specially RAM), and it | |
24 | can take a while to move all state from one machine to another. Live | |
25 | migration allows the guest to continue running while the state is | |
26 | transferred. Only while the last part of the state is transferred has | |
27 | the guest to be stopped. Typically the time that the guest is | |
28 | unresponsive during live migration is the low hundred of milliseconds | |
29 | (notice that this depends on a lot of things). | |
30 | ||
31 | Types of migration | |
32 | ================== | |
33 | ||
34 | Now that we have talked about live migration, there are several ways | |
35 | to do migration: | |
36 | ||
37 | - tcp migration: do the migration using tcp sockets | |
38 | - unix migration: do the migration using unix sockets | |
39 | - exec migration: do the migration using the stdin/stdout through a process. | |
40 | - fd migration: do the migration using an file descriptor that is | |
41 | passed to QEMU. QEMU doesn't care how this file descriptor is opened. | |
42 | ||
43 | All these four migration protocols use the same infrastructure to | |
44 | save/restore state devices. This infrastructure is shared with the | |
45 | savevm/loadvm functionality. | |
46 | ||
47 | State Live Migration | |
48 | ==================== | |
49 | ||
50 | This is used for RAM and block devices. It is not yet ported to vmstate. | |
51 | <Fill more information here> | |
52 | ||
53 | Common infrastructure | |
54 | ===================== | |
55 | ||
56 | The files, sockets or fd's that carry the migration stream are abstracted by | |
57 | the ``QEMUFile`` type (see `migration/qemu-file.h`). In most cases this | |
58 | is connected to a subtype of ``QIOChannel`` (see `io/`). | |
59 | ||
60 | Saving the state of one device | |
61 | ============================== | |
62 | ||
63 | The state of a device is saved using intermediate buffers. There are | |
64 | some helper functions to assist this saving. | |
65 | ||
66 | There is a new concept that we have to explain here: device state | |
67 | version. When we migrate a device, we save/load the state as a series | |
68 | of fields. Some times, due to bugs or new functionality, we need to | |
69 | change the state to store more/different information. We use the | |
70 | version to identify each time that we do a change. Each version is | |
71 | associated with a series of fields saved. The `save_state` always saves | |
72 | the state as the newer version. But `load_state` sometimes is able to | |
73 | load state from an older version. | |
74 | ||
75 | Legacy way | |
76 | ---------- | |
77 | ||
78 | This way is going to disappear as soon as all current users are ported to VMSTATE. | |
79 | ||
80 | Each device has to register two functions, one to save the state and | |
81 | another to load the state back. | |
82 | ||
83 | .. code:: c | |
84 | ||
85 | int register_savevm(DeviceState *dev, | |
86 | const char *idstr, | |
87 | int instance_id, | |
88 | int version_id, | |
89 | SaveStateHandler *save_state, | |
90 | LoadStateHandler *load_state, | |
91 | void *opaque); | |
92 | ||
93 | typedef void SaveStateHandler(QEMUFile *f, void *opaque); | |
94 | typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id); | |
95 | ||
96 | The important functions for the device state format are the `save_state` | |
97 | and `load_state`. Notice that `load_state` receives a version_id | |
98 | parameter to know what state format is receiving. `save_state` doesn't | |
99 | have a version_id parameter because it always uses the latest version. | |
100 | ||
101 | VMState | |
102 | ------- | |
103 | ||
104 | The legacy way of saving/loading state of the device had the problem | |
105 | that we have to maintain two functions in sync. If we did one change | |
106 | in one of them and not in the other, we would get a failed migration. | |
107 | ||
108 | VMState changed the way that state is saved/loaded. Instead of using | |
109 | a function to save the state and another to load it, it was changed to | |
110 | a declarative way of what the state consisted of. Now VMState is able | |
111 | to interpret that definition to be able to load/save the state. As | |
112 | the state is declared only once, it can't go out of sync in the | |
113 | save/load functions. | |
114 | ||
115 | An example (from hw/input/pckbd.c) | |
116 | ||
117 | .. code:: c | |
118 | ||
119 | static const VMStateDescription vmstate_kbd = { | |
120 | .name = "pckbd", | |
121 | .version_id = 3, | |
122 | .minimum_version_id = 3, | |
123 | .fields = (VMStateField[]) { | |
124 | VMSTATE_UINT8(write_cmd, KBDState), | |
125 | VMSTATE_UINT8(status, KBDState), | |
126 | VMSTATE_UINT8(mode, KBDState), | |
127 | VMSTATE_UINT8(pending, KBDState), | |
128 | VMSTATE_END_OF_LIST() | |
129 | } | |
130 | }; | |
131 | ||
132 | We are declaring the state with name "pckbd". | |
133 | The `version_id` is 3, and the fields are 4 uint8_t in a KBDState structure. | |
134 | We registered this with: | |
135 | ||
136 | .. code:: c | |
137 | ||
138 | vmstate_register(NULL, 0, &vmstate_kbd, s); | |
139 | ||
140 | Note: talk about how vmstate <-> qdev interact, and what the instance ids mean. | |
141 | ||
142 | You can search for ``VMSTATE_*`` macros for lots of types used in QEMU in | |
143 | include/hw/hw.h. | |
144 | ||
145 | More about versions | |
146 | ------------------- | |
147 | ||
148 | Version numbers are intended for major incompatible changes to the | |
149 | migration of a device, and using them breaks backwards-migration | |
150 | compatibility; in general most changes can be made by adding Subsections | |
151 | (see below) or _TEST macros (see below) which won't break compatibility. | |
152 | ||
153 | You can see that there are several version fields: | |
154 | ||
155 | - `version_id`: the maximum version_id supported by VMState for that device. | |
156 | - `minimum_version_id`: the minimum version_id that VMState is able to understand | |
157 | for that device. | |
158 | - `minimum_version_id_old`: For devices that were not able to port to vmstate, we can | |
159 | assign a function that knows how to read this old state. This field is | |
160 | ignored if there is no `load_state_old` handler. | |
161 | ||
162 | So, VMState is able to read versions from minimum_version_id to | |
163 | version_id. And the function ``load_state_old()`` (if present) is able to | |
164 | load state from minimum_version_id_old to minimum_version_id. This | |
165 | function is deprecated and will be removed when no more users are left. | |
166 | ||
167 | Saving state will always create a section with the 'version_id' value | |
168 | and thus can't be loaded by any older QEMU. | |
169 | ||
170 | Massaging functions | |
171 | ------------------- | |
172 | ||
173 | Sometimes, it is not enough to be able to save the state directly | |
174 | from one structure, we need to fill the correct values there. One | |
175 | example is when we are using kvm. Before saving the cpu state, we | |
176 | need to ask kvm to copy to QEMU the state that it is using. And the | |
177 | opposite when we are loading the state, we need a way to tell kvm to | |
178 | load the state for the cpu that we have just loaded from the QEMUFile. | |
179 | ||
180 | The functions to do that are inside a vmstate definition, and are called: | |
181 | ||
182 | - ``int (*pre_load)(void *opaque);`` | |
183 | ||
184 | This function is called before we load the state of one device. | |
185 | ||
186 | - ``int (*post_load)(void *opaque, int version_id);`` | |
187 | ||
188 | This function is called after we load the state of one device. | |
189 | ||
190 | - ``int (*pre_save)(void *opaque);`` | |
191 | ||
192 | This function is called before we save the state of one device. | |
193 | ||
194 | Example: You can look at hpet.c, that uses the three function to | |
195 | massage the state that is transferred. | |
196 | ||
197 | If you use memory API functions that update memory layout outside | |
198 | initialization (i.e., in response to a guest action), this is a strong | |
199 | indication that you need to call these functions in a `post_load` callback. | |
200 | Examples of such memory API functions are: | |
201 | ||
202 | - memory_region_add_subregion() | |
203 | - memory_region_del_subregion() | |
204 | - memory_region_set_readonly() | |
205 | - memory_region_set_enabled() | |
206 | - memory_region_set_address() | |
207 | - memory_region_set_alias_offset() | |
208 | ||
209 | Subsections | |
210 | ----------- | |
211 | ||
212 | The use of version_id allows to be able to migrate from older versions | |
213 | to newer versions of a device. But not the other way around. This | |
214 | makes very complicated to fix bugs in stable branches. If we need to | |
215 | add anything to the state to fix a bug, we have to disable migration | |
216 | to older versions that don't have that bug-fix (i.e. a new field). | |
217 | ||
218 | But sometimes, that bug-fix is only needed sometimes, not always. For | |
219 | instance, if the device is in the middle of a DMA operation, it is | |
220 | using a specific functionality, .... | |
221 | ||
222 | It is impossible to create a way to make migration from any version to | |
223 | any other version to work. But we can do better than only allowing | |
224 | migration from older versions to newer ones. For that fields that are | |
225 | only needed sometimes, we add the idea of subsections. A subsection | |
226 | is "like" a device vmstate, but with a particularity, it has a Boolean | |
227 | function that tells if that values are needed to be sent or not. If | |
228 | this functions returns false, the subsection is not sent. | |
229 | ||
230 | On the receiving side, if we found a subsection for a device that we | |
231 | don't understand, we just fail the migration. If we understand all | |
232 | the subsections, then we load the state with success. | |
233 | ||
234 | One important note is that the post_load() function is called "after" | |
235 | loading all subsections, because a newer subsection could change same | |
236 | value that it uses. | |
237 | ||
238 | Example: | |
239 | ||
240 | .. code:: c | |
241 | ||
242 | static bool ide_drive_pio_state_needed(void *opaque) | |
243 | { | |
244 | IDEState *s = opaque; | |
245 | ||
246 | return ((s->status & DRQ_STAT) != 0) | |
247 | || (s->bus->error_status & BM_STATUS_PIO_RETRY); | |
248 | } | |
249 | ||
250 | const VMStateDescription vmstate_ide_drive_pio_state = { | |
251 | .name = "ide_drive/pio_state", | |
252 | .version_id = 1, | |
253 | .minimum_version_id = 1, | |
254 | .pre_save = ide_drive_pio_pre_save, | |
255 | .post_load = ide_drive_pio_post_load, | |
256 | .needed = ide_drive_pio_state_needed, | |
257 | .fields = (VMStateField[]) { | |
258 | VMSTATE_INT32(req_nb_sectors, IDEState), | |
259 | VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1, | |
260 | vmstate_info_uint8, uint8_t), | |
261 | VMSTATE_INT32(cur_io_buffer_offset, IDEState), | |
262 | VMSTATE_INT32(cur_io_buffer_len, IDEState), | |
263 | VMSTATE_UINT8(end_transfer_fn_idx, IDEState), | |
264 | VMSTATE_INT32(elementary_transfer_size, IDEState), | |
265 | VMSTATE_INT32(packet_transfer_size, IDEState), | |
266 | VMSTATE_END_OF_LIST() | |
267 | } | |
268 | }; | |
269 | ||
270 | const VMStateDescription vmstate_ide_drive = { | |
271 | .name = "ide_drive", | |
272 | .version_id = 3, | |
273 | .minimum_version_id = 0, | |
274 | .post_load = ide_drive_post_load, | |
275 | .fields = (VMStateField[]) { | |
276 | .... several fields .... | |
277 | VMSTATE_END_OF_LIST() | |
278 | }, | |
279 | .subsections = (const VMStateDescription*[]) { | |
280 | &vmstate_ide_drive_pio_state, | |
281 | NULL | |
282 | } | |
283 | }; | |
284 | ||
285 | Here we have a subsection for the pio state. We only need to | |
286 | save/send this state when we are in the middle of a pio operation | |
287 | (that is what ``ide_drive_pio_state_needed()`` checks). If DRQ_STAT is | |
288 | not enabled, the values on that fields are garbage and don't need to | |
289 | be sent. | |
290 | ||
291 | Using a condition function that checks a 'property' to determine whether | |
292 | to send a subsection allows backwards migration compatibility when | |
293 | new subsections are added. | |
294 | ||
295 | For example: | |
296 | ||
297 | a) Add a new property using ``DEFINE_PROP_BOOL`` - e.g. support-foo and | |
298 | default it to true. | |
299 | b) Add an entry to the ``HW_COMPAT_`` for the previous version that sets | |
300 | the property to false. | |
301 | c) Add a static bool support_foo function that tests the property. | |
302 | d) Add a subsection with a .needed set to the support_foo function | |
303 | e) (potentially) Add a pre_load that sets up a default value for 'foo' | |
304 | to be used if the subsection isn't loaded. | |
305 | ||
306 | Now that subsection will not be generated when using an older | |
307 | machine type and the migration stream will be accepted by older | |
308 | QEMU versions. pre-load functions can be used to initialise state | |
309 | on the newer version so that they default to suitable values | |
310 | when loading streams created by older QEMU versions that do not | |
311 | generate the subsection. | |
312 | ||
313 | In some cases subsections are added for data that had been accidentally | |
314 | omitted by earlier versions; if the missing data causes the migration | |
315 | process to succeed but the guest to behave badly then it may be better | |
316 | to send the subsection and cause the migration to explicitly fail | |
317 | with the unknown subsection error. If the bad behaviour only happens | |
318 | with certain data values, making the subsection conditional on | |
319 | the data value (rather than the machine type) allows migrations to succeed | |
320 | in most cases. In general the preference is to tie the subsection to | |
321 | the machine type, and allow reliable migrations, unless the behaviour | |
322 | from omission of the subsection is really bad. | |
323 | ||
324 | Not sending existing elements | |
325 | ----------------------------- | |
326 | ||
327 | Sometimes members of the VMState are no longer needed: | |
328 | ||
329 | - removing them will break migration compatibility | |
330 | ||
331 | - making them version dependent and bumping the version will break backwards migration compatibility. | |
332 | ||
333 | The best way is to: | |
334 | ||
335 | a) Add a new property/compatibility/function in the same way for subsections above. | |
336 | b) replace the VMSTATE macro with the _TEST version of the macro, e.g.: | |
337 | ||
338 | ``VMSTATE_UINT32(foo, barstruct)`` | |
339 | ||
340 | becomes | |
341 | ||
342 | ``VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)`` | |
343 | ||
344 | Sometime in the future when we no longer care about the ancient versions these can be killed off. | |
345 | ||
346 | Return path | |
347 | ----------- | |
348 | ||
349 | In most migration scenarios there is only a single data path that runs | |
350 | from the source VM to the destination, typically along a single fd (although | |
351 | possibly with another fd or similar for some fast way of throwing pages across). | |
352 | ||
353 | However, some uses need two way communication; in particular the Postcopy | |
354 | destination needs to be able to request pages on demand from the source. | |
355 | ||
356 | For these scenarios there is a 'return path' from the destination to the source; | |
357 | ``qemu_file_get_return_path(QEMUFile* fwdpath)`` gives the QEMUFile* for the return | |
358 | path. | |
359 | ||
360 | Source side | |
361 | ||
362 | Forward path - written by migration thread | |
363 | Return path - opened by main thread, read by return-path thread | |
364 | ||
365 | Destination side | |
366 | ||
367 | Forward path - read by main thread | |
368 | Return path - opened by main thread, written by main thread AND postcopy | |
369 | thread (protected by rp_mutex) | |
370 | ||
371 | Postcopy | |
372 | ======== | |
373 | ||
374 | 'Postcopy' migration is a way to deal with migrations that refuse to converge | |
375 | (or take too long to converge) its plus side is that there is an upper bound on | |
376 | the amount of migration traffic and time it takes, the down side is that during | |
377 | the postcopy phase, a failure of *either* side or the network connection causes | |
378 | the guest to be lost. | |
379 | ||
380 | In postcopy the destination CPUs are started before all the memory has been | |
381 | transferred, and accesses to pages that are yet to be transferred cause | |
382 | a fault that's translated by QEMU into a request to the source QEMU. | |
383 | ||
384 | Postcopy can be combined with precopy (i.e. normal migration) so that if precopy | |
385 | doesn't finish in a given time the switch is made to postcopy. | |
386 | ||
387 | Enabling postcopy | |
388 | ----------------- | |
389 | ||
390 | To enable postcopy, issue this command on the monitor prior to the | |
391 | start of migration: | |
392 | ||
393 | ``migrate_set_capability postcopy-ram on`` | |
394 | ||
395 | The normal commands are then used to start a migration, which is still | |
396 | started in precopy mode. Issuing: | |
397 | ||
398 | ``migrate_start_postcopy`` | |
399 | ||
400 | will now cause the transition from precopy to postcopy. | |
401 | It can be issued immediately after migration is started or any | |
402 | time later on. Issuing it after the end of a migration is harmless. | |
403 | ||
404 | .. note:: | |
405 | During the postcopy phase, the bandwidth limits set using | |
406 | ``migrate_set_speed`` is ignored (to avoid delaying requested pages that | |
407 | the destination is waiting for). | |
408 | ||
409 | Postcopy device transfer | |
410 | ------------------------ | |
411 | ||
412 | Loading of device data may cause the device emulation to access guest RAM | |
413 | that may trigger faults that have to be resolved by the source, as such | |
414 | the migration stream has to be able to respond with page data *during* the | |
415 | device load, and hence the device data has to be read from the stream completely | |
416 | before the device load begins to free the stream up. This is achieved by | |
417 | 'packaging' the device data into a blob that's read in one go. | |
418 | ||
419 | Source behaviour | |
420 | ---------------- | |
421 | ||
422 | Until postcopy is entered the migration stream is identical to normal | |
423 | precopy, except for the addition of a 'postcopy advise' command at | |
424 | the beginning, to tell the destination that postcopy might happen. | |
425 | When postcopy starts the source sends the page discard data and then | |
426 | forms the 'package' containing: | |
427 | ||
428 | - Command: 'postcopy listen' | |
429 | - The device state | |
430 | ||
431 | A series of sections, identical to the precopy streams device state stream | |
432 | containing everything except postcopiable devices (i.e. RAM) | |
433 | - Command: 'postcopy run' | |
434 | ||
435 | The 'package' is sent as the data part of a Command: ``CMD_PACKAGED``, and the | |
436 | contents are formatted in the same way as the main migration stream. | |
437 | ||
438 | During postcopy the source scans the list of dirty pages and sends them | |
439 | to the destination without being requested (in much the same way as precopy), | |
440 | however when a page request is received from the destination, the dirty page | |
441 | scanning restarts from the requested location. This causes requested pages | |
442 | to be sent quickly, and also causes pages directly after the requested page | |
443 | to be sent quickly in the hope that those pages are likely to be used | |
444 | by the destination soon. | |
445 | ||
446 | Destination behaviour | |
447 | --------------------- | |
448 | ||
449 | Initially the destination looks the same as precopy, with a single thread | |
450 | reading the migration stream; the 'postcopy advise' and 'discard' commands | |
451 | are processed to change the way RAM is managed, but don't affect the stream | |
452 | processing. | |
453 | ||
454 | :: | |
455 | ||
456 | ------------------------------------------------------------------------------ | |
457 | 1 2 3 4 5 6 7 | |
458 | main -----DISCARD-CMD_PACKAGED ( LISTEN DEVICE DEVICE DEVICE RUN ) | |
459 | thread | | | |
460 | | (page request) | |
461 | | \___ | |
462 | v \ | |
463 | listen thread: --- page -- page -- page -- page -- page -- | |
464 | ||
465 | a b c | |
466 | ------------------------------------------------------------------------------ | |
467 | ||
468 | - On receipt of ``CMD_PACKAGED`` (1) | |
469 | ||
470 | All the data associated with the package - the ( ... ) section in the diagram - | |
471 | is read into memory, and the main thread recurses into qemu_loadvm_state_main | |
472 | to process the contents of the package (2) which contains commands (3,6) and | |
473 | devices (4...) | |
474 | ||
475 | - On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package) | |
476 | ||
477 | a new thread (a) is started that takes over servicing the migration stream, | |
478 | while the main thread carries on loading the package. It loads normal | |
479 | background page data (b) but if during a device load a fault happens (5) | |
480 | the returned page (c) is loaded by the listen thread allowing the main | |
481 | threads device load to carry on. | |
482 | ||
483 | - The last thing in the ``CMD_PACKAGED`` is a 'RUN' command (6) | |
484 | ||
485 | letting the destination CPUs start running. At the end of the | |
486 | ``CMD_PACKAGED`` (7) the main thread returns to normal running behaviour and | |
487 | is no longer used by migration, while the listen thread carries on servicing | |
488 | page data until the end of migration. | |
489 | ||
490 | Postcopy states | |
491 | --------------- | |
492 | ||
493 | Postcopy moves through a series of states (see postcopy_state) from | |
494 | ADVISE->DISCARD->LISTEN->RUNNING->END | |
495 | ||
496 | - Advise | |
497 | ||
498 | Set at the start of migration if postcopy is enabled, even | |
499 | if it hasn't had the start command; here the destination | |
500 | checks that its OS has the support needed for postcopy, and performs | |
501 | setup to ensure the RAM mappings are suitable for later postcopy. | |
502 | The destination will fail early in migration at this point if the | |
503 | required OS support is not present. | |
504 | (Triggered by reception of POSTCOPY_ADVISE command) | |
505 | ||
506 | - Discard | |
507 | ||
508 | Entered on receipt of the first 'discard' command; prior to | |
509 | the first Discard being performed, hugepages are switched off | |
510 | (using madvise) to ensure that no new huge pages are created | |
511 | during the postcopy phase, and to cause any huge pages that | |
512 | have discards on them to be broken. | |
513 | ||
514 | - Listen | |
515 | ||
516 | The first command in the package, POSTCOPY_LISTEN, switches | |
517 | the destination state to Listen, and starts a new thread | |
518 | (the 'listen thread') which takes over the job of receiving | |
519 | pages off the migration stream, while the main thread carries | |
520 | on processing the blob. With this thread able to process page | |
521 | reception, the destination now 'sensitises' the RAM to detect | |
522 | any access to missing pages (on Linux using the 'userfault' | |
523 | system). | |
524 | ||
525 | - Running | |
526 | ||
527 | POSTCOPY_RUN causes the destination to synchronise all | |
528 | state and start the CPUs and IO devices running. The main | |
529 | thread now finishes processing the migration package and | |
530 | now carries on as it would for normal precopy migration | |
531 | (although it can't do the cleanup it would do as it | |
532 | finishes a normal migration). | |
533 | ||
534 | - End | |
535 | ||
536 | The listen thread can now quit, and perform the cleanup of migration | |
537 | state, the migration is now complete. | |
538 | ||
539 | Source side page maps | |
540 | --------------------- | |
541 | ||
542 | The source side keeps two bitmaps during postcopy; 'the migration bitmap' | |
543 | and 'unsent map'. The 'migration bitmap' is basically the same as in | |
544 | the precopy case, and holds a bit to indicate that page is 'dirty' - | |
545 | i.e. needs sending. During the precopy phase this is updated as the CPU | |
546 | dirties pages, however during postcopy the CPUs are stopped and nothing | |
547 | should dirty anything any more. | |
548 | ||
549 | The 'unsent map' is used for the transition to postcopy. It is a bitmap that | |
550 | has a bit cleared whenever a page is sent to the destination, however during | |
551 | the transition to postcopy mode it is combined with the migration bitmap | |
552 | to form a set of pages that: | |
553 | ||
554 | a) Have been sent but then redirtied (which must be discarded) | |
555 | b) Have not yet been sent - which also must be discarded to cause any | |
556 | transparent huge pages built during precopy to be broken. | |
557 | ||
558 | Note that the contents of the unsentmap are sacrificed during the calculation | |
559 | of the discard set and thus aren't valid once in postcopy. The dirtymap | |
560 | is still valid and is used to ensure that no page is sent more than once. Any | |
561 | request for a page that has already been sent is ignored. Duplicate requests | |
562 | such as this can happen as a page is sent at about the same time the | |
563 | destination accesses it. | |
564 | ||
565 | Postcopy with hugepages | |
566 | ----------------------- | |
567 | ||
568 | Postcopy now works with hugetlbfs backed memory: | |
569 | ||
570 | a) The linux kernel on the destination must support userfault on hugepages. | |
571 | b) The huge-page configuration on the source and destination VMs must be | |
572 | identical; i.e. RAMBlocks on both sides must use the same page size. | |
573 | c) Note that ``-mem-path /dev/hugepages`` will fall back to allocating normal | |
574 | RAM if it doesn't have enough hugepages, triggering (b) to fail. | |
575 | Using ``-mem-prealloc`` enforces the allocation using hugepages. | |
576 | d) Care should be taken with the size of hugepage used; postcopy with 2MB | |
577 | hugepages works well, however 1GB hugepages are likely to be problematic | |
578 | since it takes ~1 second to transfer a 1GB hugepage across a 10Gbps link, | |
579 | and until the full page is transferred the destination thread is blocked. |