1 // -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
2 // vim: ts=8 sw=2 smarttab
4 * Ceph - scalable distributed file system
6 * Copyright (C) 2004-2006 Sage Weil <sage@newdream.net>
8 * This is free software; you can redistribute it and/or
9 * modify it under the terms of the GNU Lesser General Public
10 * License version 2.1, as published by the Free Software
11 * Foundation. See file COPYING.
16 #include "common/debug.h"
17 #include "common/errno.h"
18 #include "common/safe_io.h"
19 #include "FileJournal.h"
20 #include "include/color.h"
21 #include "common/perf_counters.h"
22 #include "FileStore.h"
24 #include "include/compat.h"
31 #include <sys/types.h>
33 #include <sys/mount.h>
35 #include "common/blkdev.h"
36 #if defined(__linux__)
37 #include "common/linux_version.h"
40 #if defined(__FreeBSD__)
41 #define O_DSYNC O_SYNC
44 #define dout_context cct
45 #define dout_subsys ceph_subsys_journal
47 #define dout_prefix *_dout << "journal "
49 const static int64_t ONE_MEG(1 << 20);
50 const static int CEPH_DIRECTIO_ALIGNMENT(4096);
53 int FileJournal::_open(bool forwrite
, bool create
)
60 flags
|= O_DIRECT
| O_DSYNC
;
68 if (TEMP_FAILURE_RETRY(::close(fd
))) {
70 derr
<< "FileJournal::_open: error closing old fd: "
71 << cpp_strerror(err
) << dendl
;
74 fd
= TEMP_FAILURE_RETRY(::open(fn
.c_str(), flags
, 0644));
77 dout(2) << "FileJournal::_open unable to open journal "
78 << fn
<< ": " << cpp_strerror(err
) << dendl
;
83 ret
= ::fstat(fd
, &st
);
86 derr
<< "FileJournal::_open: unable to fstat journal: " << cpp_strerror(ret
) << dendl
;
91 if (S_ISBLK(st
.st_mode
)) {
92 ret
= _open_block_device();
93 } else if (S_ISREG(st
.st_mode
)) {
94 if (aio
&& !force_aio
) {
95 derr
<< "FileJournal::_open: disabling aio for non-block journal. Use "
96 << "journal_force_aio to force use of aio anyway" << dendl
;
99 ret
= _open_file(st
.st_size
, st
.st_blksize
, create
);
101 derr
<< "FileJournal::_open: wrong journal file type: " << st
.st_mode
112 ret
= io_setup(128, &aio_ctx
);
115 // Contrary to naive expectations -EAGIAN means ...
117 derr
<< "FileJournal::_open: user's limit of aio events exceeded. "
118 << "Try increasing /proc/sys/fs/aio-max-nr" << dendl
;
121 derr
<< "FileJournal::_open: unable to setup io_context " << cpp_strerror(-ret
) << dendl
;
129 /* We really want max_size to be a multiple of block_size. */
130 max_size
-= max_size
% block_size
;
132 dout(1) << "_open " << fn
<< " fd " << fd
134 << " bytes, block size " << block_size
135 << " bytes, directio = " << directio
141 VOID_TEMP_FAILURE_RETRY(::close(fd
));
146 int FileJournal::_open_block_device()
149 int ret
= get_block_device_size(fd
, &bdev_sz
);
151 dout(0) << __func__
<< ": failed to read block device size." << dendl
;
155 /* Check for bdev_sz too small */
156 if (bdev_sz
< ONE_MEG
) {
157 dout(0) << __func__
<< ": your block device must be at least "
158 << ONE_MEG
<< " bytes to be used for a Ceph journal." << dendl
;
162 dout(10) << __func__
<< ": ignoring osd journal size. "
163 << "We'll use the entire block device (size: " << bdev_sz
<< ")"
167 block_size
= cct
->_conf
->journal_block_size
;
169 if (cct
->_conf
->journal_discard
) {
170 discard
= block_device_support_discard(fn
.c_str());
171 dout(10) << fn
<< " support discard: " << (int)discard
<< dendl
;
177 int FileJournal::_open_file(int64_t oldsize
, blksize_t blksize
,
181 int64_t conf_journal_sz(cct
->_conf
->osd_journal_size
);
182 conf_journal_sz
<<= 20;
184 if ((cct
->_conf
->osd_journal_size
== 0) && (oldsize
< ONE_MEG
)) {
185 derr
<< "I'm sorry, I don't know how large of a journal to create."
186 << "Please specify a block device to use as the journal OR "
187 << "set osd_journal_size in your ceph.conf" << dendl
;
191 if (create
&& (oldsize
< conf_journal_sz
)) {
192 uint64_t newsize(conf_journal_sz
);
193 dout(10) << __func__
<< " _open extending to " << newsize
<< " bytes" << dendl
;
194 ret
= ::ftruncate(fd
, newsize
);
197 derr
<< "FileJournal::_open_file : unable to extend journal to "
198 << newsize
<< " bytes: " << cpp_strerror(err
) << dendl
;
201 #ifdef HAVE_POSIX_FALLOCATE
202 ret
= ::posix_fallocate(fd
, 0, newsize
);
204 derr
<< "FileJournal::_open_file : unable to preallocation journal to "
205 << newsize
<< " bytes: " << cpp_strerror(ret
) << dendl
;
209 #elif defined(__APPLE__)
211 store
.fst_flags
= F_ALLOCATECONTIG
;
212 store
.fst_posmode
= F_PEOFPOSMODE
;
213 store
.fst_offset
= 0;
214 store
.fst_length
= newsize
;
216 ret
= ::fcntl(fd
, F_PREALLOCATE
, &store
);
219 derr
<< "FileJournal::_open_file : unable to preallocation journal to "
220 << newsize
<< " bytes: " << cpp_strerror(ret
) << dendl
;
225 # error "Journal pre-allocation not supported on platform."
231 block_size
= cct
->_conf
->journal_block_size
;
233 if (create
&& cct
->_conf
->journal_zero_on_create
) {
234 derr
<< "FileJournal::_open_file : zeroing journal" << dendl
;
235 uint64_t write_size
= 1 << 20;
237 ret
= ::posix_memalign((void **)&buf
, block_size
, write_size
);
241 memset(static_cast<void*>(buf
), 0, write_size
);
243 for (; (i
+ write_size
) <= (uint64_t)max_size
; i
+= write_size
) {
244 ret
= ::pwrite(fd
, static_cast<void*>(buf
), write_size
, i
);
250 if (i
< (uint64_t)max_size
) {
251 ret
= ::pwrite(fd
, static_cast<void*>(buf
), max_size
- i
, i
);
261 dout(10) << "_open journal is not a block device, NOT checking disk "
262 << "write cache on '" << fn
<< "'" << dendl
;
267 // This can not be used on an active journal
268 int FileJournal::check()
273 ret
= _open(false, false);
277 ret
= read_header(&header
);
281 if (header
.fsid
!= fsid
) {
282 derr
<< "check: ondisk fsid " << header
.fsid
<< " doesn't match expected " << fsid
283 << ", invalid (someone else's?) journal" << dendl
;
288 dout(1) << "check: header looks ok" << dendl
;
297 int FileJournal::create()
300 int64_t needed_space
;
303 dout(2) << "create " << fn
<< " fsid " << fsid
<< dendl
;
305 ret
= _open(true, true);
309 // write empty header
311 header
.flags
= header_t::FLAG_CRC
; // enable crcs on any new journal.
313 header
.max_size
= max_size
;
314 header
.block_size
= block_size
;
315 if (cct
->_conf
->journal_block_align
|| directio
)
316 header
.alignment
= block_size
;
318 header
.alignment
= 16; // at least stay word aligned on 64bit machines...
320 header
.start
= get_top();
321 header
.start_seq
= 0;
323 print_header(header
);
325 // static zeroed buffer for alignment padding
327 zero_buf
= new char[header
.alignment
];
328 memset(zero_buf
, 0, header
.alignment
);
330 bp
= prepare_header();
331 if (TEMP_FAILURE_RETRY(::pwrite(fd
, bp
.c_str(), bp
.length(), 0)) < 0) {
333 derr
<< "FileJournal::create : create write header error "
334 << cpp_strerror(ret
) << dendl
;
338 // zero first little bit, too.
339 ret
= posix_memalign(&buf
, block_size
, block_size
);
342 derr
<< "FileJournal::create: failed to allocate " << block_size
343 << " bytes of memory: " << cpp_strerror(ret
) << dendl
;
346 memset(buf
, 0, block_size
);
347 if (TEMP_FAILURE_RETRY(::pwrite(fd
, buf
, block_size
, get_top())) < 0) {
349 derr
<< "FileJournal::create: error zeroing first " << block_size
350 << " bytes " << cpp_strerror(ret
) << dendl
;
354 needed_space
= ((int64_t)cct
->_conf
->osd_max_write_size
) << 20;
355 needed_space
+= (2 * sizeof(entry_header_t
)) + get_top();
356 if (header
.max_size
- header
.start
< needed_space
) {
357 derr
<< "FileJournal::create: OSD journal is not large enough to hold "
358 << "osd_max_write_size bytes!" << dendl
;
363 dout(2) << "create done" << dendl
;
370 if (TEMP_FAILURE_RETRY(::close(fd
)) < 0) {
372 derr
<< "FileJournal::create: error closing fd: " << cpp_strerror(ret
)
380 // This can not be used on an active journal
381 int FileJournal::peek_fsid(uuid_d
& fsid
)
384 int r
= _open(false, false);
387 r
= read_header(&header
);
396 int FileJournal::open(uint64_t fs_op_seq
)
398 dout(2) << "open " << fn
<< " fsid " << fsid
<< " fs_op_seq " << fs_op_seq
<< dendl
;
400 uint64_t next_seq
= fs_op_seq
+ 1;
403 int err
= _open(false);
407 // assume writeable, unless...
409 write_pos
= get_top();
412 err
= read_header(&header
);
416 // static zeroed buffer for alignment padding
418 zero_buf
= new char[header
.alignment
];
419 memset(zero_buf
, 0, header
.alignment
);
421 dout(10) << "open header.fsid = " << header
.fsid
422 //<< " vs expected fsid = " << fsid
424 if (header
.fsid
!= fsid
) {
425 derr
<< "FileJournal::open: ondisk fsid " << header
.fsid
<< " doesn't match expected " << fsid
426 << ", invalid (someone else's?) journal" << dendl
;
430 if (header
.max_size
> max_size
) {
431 dout(2) << "open journal size " << header
.max_size
<< " > current " << max_size
<< dendl
;
435 if (header
.block_size
!= block_size
) {
436 dout(2) << "open journal block size " << header
.block_size
<< " != current " << block_size
<< dendl
;
440 if (header
.max_size
% header
.block_size
) {
441 dout(2) << "open journal max size " << header
.max_size
442 << " not a multiple of block size " << header
.block_size
<< dendl
;
446 if (header
.alignment
!= block_size
&& directio
) {
447 dout(0) << "open journal alignment " << header
.alignment
<< " does not match block size "
448 << block_size
<< " (required for direct_io journal mode)" << dendl
;
452 if ((header
.alignment
% CEPH_DIRECTIO_ALIGNMENT
) && directio
) {
453 dout(0) << "open journal alignment " << header
.alignment
454 << " is not multiple of minimum directio alignment "
455 << CEPH_DIRECTIO_ALIGNMENT
<< " (required for direct_io journal mode)"
461 // looks like a valid header.
462 write_pos
= 0; // not writeable yet
464 journaled_seq
= header
.committed_up_to
;
467 read_pos
= header
.start
;
468 seq
= header
.start_seq
;
472 off64_t old_pos
= read_pos
;
473 if (!read_entry(bl
, seq
)) {
474 dout(10) << "open reached end of journal." << dendl
;
477 if (seq
> next_seq
) {
478 dout(10) << "open entry " << seq
<< " len " << bl
.length() << " > next_seq " << next_seq
479 << ", ignoring journal contents"
482 last_committed_seq
= 0;
485 if (seq
== next_seq
) {
486 dout(10) << "open reached seq " << seq
<< dendl
;
490 seq
++; // next event should follow.
499 void FileJournal::_close(int fd
) const
501 VOID_TEMP_FAILURE_RETRY(::close(fd
));
504 void FileJournal::close()
506 dout(1) << "close " << fn
<< dendl
;
508 // stop writer thread
512 assert(writeq_empty());
513 assert(!must_write_header
);
520 int FileJournal::dump(ostream
& out
)
522 return _dump(out
, false);
525 int FileJournal::simple_dump(ostream
& out
)
527 return _dump(out
, true);
530 int FileJournal::_dump(ostream
& out
, bool simple
)
532 JSONFormatter
f(true);
533 int ret
= _fdump(f
, simple
);
538 int FileJournal::_fdump(Formatter
&f
, bool simple
)
540 dout(10) << "_fdump" << dendl
;
543 int err
= _open(false, false);
547 err
= read_header(&header
);
553 off64_t next_pos
= header
.start
;
555 f
.open_object_section("journal");
557 f
.open_object_section("header");
558 f
.dump_unsigned("flags", header
.flags
);
561 f
.dump_string("fsid", os
.str());
562 f
.dump_unsigned("block_size", header
.block_size
);
563 f
.dump_unsigned("alignment", header
.alignment
);
564 f
.dump_int("max_size", header
.max_size
);
565 f
.dump_int("start", header
.start
);
566 f
.dump_unsigned("committed_up_to", header
.committed_up_to
);
567 f
.dump_unsigned("start_seq", header
.start_seq
);
570 f
.open_array_section("entries");
571 uint64_t seq
= header
.start_seq
;
574 off64_t pos
= next_pos
;
577 dout(2) << "_dump -- not readable" << dendl
;
582 read_entry_result result
= do_read_entry(
588 if (result
!= SUCCESS
) {
589 if (seq
< header
.committed_up_to
) {
590 dout(2) << "Unable to read past sequence " << seq
591 << " but header indicates the journal has committed up through "
592 << header
.committed_up_to
<< ", journal is corrupt" << dendl
;
595 dout(25) << ss
.str() << dendl
;
596 dout(25) << "No further valid entries found, journal is most likely valid"
601 f
.open_object_section("entry");
602 f
.dump_unsigned("offset", pos
);
603 f
.dump_unsigned("seq", seq
);
605 f
.dump_unsigned("bl.length", bl
.length());
607 f
.open_array_section("transactions");
608 bufferlist::iterator p
= bl
.begin();
611 ObjectStore::Transaction
t(p
);
612 f
.open_object_section("transaction");
613 f
.dump_unsigned("trans_num", trans_num
);
625 dout(10) << "dump finish" << dendl
;
632 void FileJournal::start_writer()
636 write_thread
.create("journal_write");
639 write_finish_thread
.create("journal_wrt_fin");
643 void FileJournal::stop_writer()
645 // Do nothing if writer already stopped or never started
649 Mutex::Locker
l(write_lock
);
650 Mutex::Locker
p(writeq_lock
);
652 writeq_cond
.Signal();
653 // Doesn't hurt to signal commit_cond in case thread is waiting there
654 // and caller didn't use committed_thru() first.
655 commit_cond
.Signal();
659 // write journal header now so that we have less to replay on remount
664 // stop aio completeion thread *after* writer thread has stopped
665 // and has submitted all of its io
666 if (aio
&& !aio_stop
) {
670 write_finish_cond
.Signal();
672 write_finish_thread
.join();
679 void FileJournal::print_header(const header_t
&header
) const
681 dout(10) << "header: block_size " << header
.block_size
682 << " alignment " << header
.alignment
683 << " max_size " << header
.max_size
685 dout(10) << "header: start " << header
.start
<< dendl
;
686 dout(10) << " write_pos " << write_pos
<< dendl
;
689 int FileJournal::read_header(header_t
*hdr
) const
691 dout(10) << "read_header" << dendl
;
694 buffer::ptr bp
= buffer::create_page_aligned(block_size
);
695 char* bpdata
= bp
.c_str();
696 int r
= ::pread(fd
, bpdata
, bp
.length(), 0);
700 dout(0) << "read_header got " << cpp_strerror(err
) << dendl
;
704 // don't use bp.zero() here, because it also invalidates
705 // crc cache (which is not yet populated anyway)
706 if (bp
.length() != (size_t)r
) {
707 // r will be always less or equal than bp.length
709 memset(bpdata
, 0, bp
.length() - r
);
712 bl
.push_back(std::move(bp
));
715 bufferlist::iterator p
= bl
.begin();
718 catch (buffer::error
& e
) {
719 derr
<< "read_header error decoding journal header" << dendl
;
725 * Unfortunately we weren't initializing the flags field for new
726 * journals! Aie. This is safe(ish) now that we have only one
727 * flag. Probably around when we add the next flag we need to
728 * remove this or else this (eventually old) code will clobber newer
731 if (hdr
->flags
> 3) {
732 derr
<< "read_header appears to have gibberish flags; assuming 0" << dendl
;
741 bufferptr
FileJournal::prepare_header()
745 Mutex::Locker
l(finisher_lock
);
746 header
.committed_up_to
= journaled_seq
;
748 ::encode(header
, bl
);
749 bufferptr bp
= buffer::create_page_aligned(get_top());
750 // don't use bp.zero() here, because it also invalidates
751 // crc cache (which is not yet populated anyway)
752 char* data
= bp
.c_str();
753 memcpy(data
, bl
.c_str(), bl
.length());
755 memset(data
, 0, bp
.length()-bl
.length());
759 void FileJournal::write_header_sync()
761 Mutex::Locker
locker(write_lock
);
762 must_write_header
= true;
765 dout(20) << __func__
<< " finish" << dendl
;
768 int FileJournal::check_for_full(uint64_t seq
, off64_t pos
, off64_t size
)
771 if (full_state
!= FULL_NOTFULL
)
774 // take 1 byte off so that we only get pos == header.start on EMPTY, never on FULL.
776 if (pos
>= header
.start
)
777 room
= (header
.max_size
- pos
) + (header
.start
- get_top()) - 1;
779 room
= header
.start
- pos
- 1;
780 dout(10) << "room " << room
<< " max_size " << max_size
<< " pos " << pos
<< " header.start " << header
.start
781 << " top " << get_top() << dendl
;
784 if (room
>= (header
.max_size
>> 1) &&
785 room
- size
< (header
.max_size
>> 1)) {
786 dout(10) << " passing half full mark, triggering commit" << dendl
;
787 do_sync_cond
->SloppySignal(); // initiate a real commit so we can trim
792 dout(10) << "check_for_full at " << pos
<< " : " << size
<< " < " << room
<< dendl
;
793 if (pos
+ size
> header
.max_size
)
794 must_write_header
= true;
799 dout(1) << "check_for_full at " << pos
<< " : JOURNAL FULL "
800 << pos
<< " >= " << room
801 << " (max_size " << header
.max_size
<< " start " << header
.start
<< ")"
804 off64_t max
= header
.max_size
- get_top();
806 dout(0) << "JOURNAL TOO SMALL: continuing, but slow: item " << size
<< " > journal " << max
<< " (usable)" << dendl
;
811 int FileJournal::prepare_multi_write(bufferlist
& bl
, uint64_t& orig_ops
, uint64_t& orig_bytes
)
813 // gather queued writes
814 off64_t queue_pos
= write_pos
;
816 int eleft
= cct
->_conf
->journal_max_write_entries
;
817 unsigned bmax
= cct
->_conf
->journal_max_write_bytes
;
819 if (full_state
!= FULL_NOTFULL
)
822 while (!writeq_empty()) {
823 list
<write_item
> items
;
824 batch_pop_write(items
);
825 list
<write_item
>::iterator it
= items
.begin();
826 while (it
!= items
.end()) {
827 uint64_t bytes
= it
->bl
.length();
828 int r
= prepare_single_write(*it
, bl
, queue_pos
, orig_ops
, orig_bytes
);
829 if (r
== 0) { // prepare ok, delete it
833 Mutex::Locker
locker(aio_lock
);
834 assert(aio_write_queue_ops
> 0);
835 aio_write_queue_ops
--;
836 assert(aio_write_queue_bytes
>= bytes
);
837 aio_write_queue_bytes
-= bytes
;
844 // the journal maybe full, insert the left item to writeq
845 batch_unpop_write(items
);
847 goto out
; // commit what we have
850 logger
->inc(l_filestore_journal_full
);
853 dout(20) << "prepare_multi_write full on first entry, need to wait" << dendl
;
855 dout(20) << "prepare_multi_write full on first entry, restarting journal" << dendl
;
857 // throw out what we have so far
858 full_state
= FULL_FULL
;
859 while (!writeq_empty()) {
860 complete_write(1, peek_write().orig_len
);
863 print_header(header
);
866 return -ENOSPC
; // hrm, full on first op
870 dout(20) << "prepare_multi_write hit max events per write "
871 << cct
->_conf
->journal_max_write_entries
<< dendl
;
872 batch_unpop_write(items
);
877 if (bl
.length() >= bmax
) {
878 dout(20) << "prepare_multi_write hit max write size "
879 << cct
->_conf
->journal_max_write_bytes
<< dendl
;
880 batch_unpop_write(items
);
888 dout(20) << "prepare_multi_write queue_pos now " << queue_pos
<< dendl
;
889 assert((write_pos
+ bl
.length() == queue_pos
) ||
890 (write_pos
+ bl
.length() - header
.max_size
+ get_top() == queue_pos
));
895 void FileJournal::queue_write_fin(uint64_t seq, Context *fin)
897 writing_seq.push_back(seq);
898 if (!waiting_for_notfull.empty()) {
899 // make sure previously unjournaled stuff waiting for UNFULL triggers
900 // _before_ newly journaled stuff does
901 dout(10) << "queue_write_fin will defer seq " << seq << " callback " << fin
902 << " until after UNFULL" << dendl;
903 C_Gather *g = new C_Gather(writeq.front().fin);
904 writing_fin.push_back(g->new_sub());
905 waiting_for_notfull.push_back(g->new_sub());
907 writing_fin.push_back(writeq.front().fin);
908 dout(20) << "queue_write_fin seq " << seq << " callback " << fin << dendl;
913 void FileJournal::queue_completions_thru(uint64_t seq
)
915 assert(finisher_lock
.is_locked());
916 utime_t now
= ceph_clock_now();
917 list
<completion_item
> items
;
918 batch_pop_completions(items
);
919 list
<completion_item
>::iterator it
= items
.begin();
920 while (it
!= items
.end()) {
921 completion_item
& next
= *it
;
926 dout(10) << "queue_completions_thru seq " << seq
927 << " queueing seq " << next
.seq
928 << " " << next
.finish
929 << " lat " << lat
<< dendl
;
931 logger
->tinc(l_filestore_journal_latency
, lat
);
934 finisher
->queue(next
.finish
);
935 if (next
.tracked_op
) {
936 next
.tracked_op
->mark_event("journaled_completion_queued");
937 next
.tracked_op
->journal_trace
.event("queued completion");
938 next
.tracked_op
->journal_trace
.keyval("completed through", seq
);
942 batch_unpop_completions(items
);
943 finisher_cond
.Signal();
947 int FileJournal::prepare_single_write(write_item
&next_write
, bufferlist
& bl
, off64_t
& queue_pos
, uint64_t& orig_ops
, uint64_t& orig_bytes
)
949 uint64_t seq
= next_write
.seq
;
950 bufferlist
&ebl
= next_write
.bl
;
951 off64_t size
= ebl
.length();
953 int r
= check_for_full(seq
, queue_pos
, size
);
955 return r
; // ENOSPC or EAGAIN
957 uint32_t orig_len
= next_write
.orig_len
;
958 orig_bytes
+= orig_len
;
961 // add to write buffer
962 dout(15) << "prepare_single_write " << orig_ops
<< " will write " << queue_pos
<< " : seq " << seq
963 << " len " << orig_len
<< " -> " << size
<< dendl
;
965 unsigned seq_offset
= offsetof(entry_header_t
, seq
);
966 unsigned magic1_offset
= offsetof(entry_header_t
, magic1
);
967 unsigned magic2_offset
= offsetof(entry_header_t
, magic2
);
969 bufferptr headerptr
= ebl
.buffers().front();
971 uint64_t _queue_pos
= queue_pos
;
972 uint64_t magic2
= entry_header_t::make_magic(seq
, orig_len
, header
.get_fsid64());
973 headerptr
.copy_in(seq_offset
, sizeof(uint64_t), (char *)&_seq
);
974 headerptr
.copy_in(magic1_offset
, sizeof(uint64_t), (char *)&_queue_pos
);
975 headerptr
.copy_in(magic2_offset
, sizeof(uint64_t), (char *)&magic2
);
977 bufferptr footerptr
= ebl
.buffers().back();
978 unsigned post_offset
= footerptr
.length() - sizeof(entry_header_t
);
979 footerptr
.copy_in(post_offset
+ seq_offset
, sizeof(uint64_t), (char *)&_seq
);
980 footerptr
.copy_in(post_offset
+ magic1_offset
, sizeof(uint64_t), (char *)&_queue_pos
);
981 footerptr
.copy_in(post_offset
+ magic2_offset
, sizeof(uint64_t), (char *)&magic2
);
983 bl
.claim_append(ebl
);
984 if (next_write
.tracked_op
) {
985 next_write
.tracked_op
->mark_event("write_thread_in_journal_buffer");
986 next_write
.tracked_op
->journal_trace
.event("prepare_single_write");
989 journalq
.push_back(pair
<uint64_t,off64_t
>(seq
, queue_pos
));
993 if (queue_pos
>= header
.max_size
)
994 queue_pos
= queue_pos
+ get_top() - header
.max_size
;
999 void FileJournal::check_align(off64_t pos
, bufferlist
& bl
)
1001 // make sure list segments are page aligned
1002 if (directio
&& !bl
.is_aligned_size_and_memory(block_size
, CEPH_DIRECTIO_ALIGNMENT
)) {
1003 assert((bl
.length() & (CEPH_DIRECTIO_ALIGNMENT
- 1)) == 0);
1004 assert((pos
& (CEPH_DIRECTIO_ALIGNMENT
- 1)) == 0);
1005 assert(0 == "bl was not aligned");
1009 int FileJournal::write_bl(off64_t
& pos
, bufferlist
& bl
)
1013 off64_t spos
= ::lseek64(fd
, pos
, SEEK_SET
);
1016 derr
<< "FileJournal::write_bl : lseek64 failed " << cpp_strerror(ret
) << dendl
;
1019 ret
= bl
.write_fd(fd
);
1021 derr
<< "FileJournal::write_bl : write_fd failed: " << cpp_strerror(ret
) << dendl
;
1025 if (pos
== header
.max_size
)
1030 void FileJournal::do_write(bufferlist
& bl
)
1033 if (bl
.length() == 0 && !must_write_header
)
1037 if (cct
->_conf
->journal_write_header_frequency
&&
1038 (((++journaled_since_start
) %
1039 cct
->_conf
->journal_write_header_frequency
) == 0)) {
1040 must_write_header
= true;
1043 if (must_write_header
) {
1044 must_write_header
= false;
1045 hbp
= prepare_header();
1048 dout(15) << "do_write writing " << write_pos
<< "~" << bl
.length()
1049 << (hbp
.length() ? " + header":"")
1052 utime_t from
= ceph_clock_now();
1055 off64_t pos
= write_pos
;
1058 write_pos
+= bl
.length();
1059 if (write_pos
>= header
.max_size
)
1060 write_pos
= write_pos
- header
.max_size
+ get_top();
1062 write_lock
.Unlock();
1066 if (pos
+ bl
.length() > header
.max_size
) {
1067 bufferlist first
, second
;
1068 split
= header
.max_size
- pos
;
1069 first
.substr_of(bl
, 0, split
);
1070 second
.substr_of(bl
, split
, bl
.length() - split
);
1071 assert(first
.length() + second
.length() == bl
.length());
1072 dout(10) << "do_write wrapping, first bit at " << pos
<< " len " << first
.length()
1073 << " second bit len " << second
.length() << " (orig len " << bl
.length() << ")" << dendl
;
1075 //Save pos to write first piece second
1076 off64_t first_pos
= pos
;
1081 // be sneaky: include the header in the second fragment
1082 second
.push_front(hbp
);
1083 pos
= 0; // we included the header
1085 // Write the second portion first possible with the header, so
1086 // do_read_entry() won't even get a valid entry_header_t if there
1087 // is a crash between the two writes.
1089 if (write_bl(pos
, second
)) {
1090 derr
<< "FileJournal::do_write: write_bl(pos=" << orig_pos
1091 << ") failed" << dendl
;
1092 check_align(pos
, second
);
1095 orig_pos
= first_pos
;
1096 if (write_bl(first_pos
, first
)) {
1097 derr
<< "FileJournal::do_write: write_bl(pos=" << orig_pos
1098 << ") failed" << dendl
;
1099 check_align(first_pos
, first
);
1102 assert(first_pos
== get_top());
1106 if (TEMP_FAILURE_RETRY(::pwrite(fd
, hbp
.c_str(), hbp
.length(), 0)) < 0) {
1108 derr
<< "FileJournal::do_write: pwrite(fd=" << fd
1109 << ", hbp.length=" << hbp
.length() << ") failed :"
1110 << cpp_strerror(err
) << dendl
;
1115 if (write_bl(pos
, bl
)) {
1116 derr
<< "FileJournal::do_write: write_bl(pos=" << pos
1117 << ") failed" << dendl
;
1118 check_align(pos
, bl
);
1124 dout(20) << "do_write fsync" << dendl
;
1127 * We'd really love to have a fsync_range or fdatasync_range and do a:
1130 * ::fsync_range(fd, header.max_size - split, split)l
1131 * ::fsync_range(fd, get_top(), bl.length() - split);
1133 * ::fsync_range(fd, write_pos, bl.length())
1135 * NetBSD and AIX apparently have it, and adding it to Linux wouldn't be
1136 * too hard given all the underlying infrastructure already exist.
1138 * NOTE: using sync_file_range here would not be safe as it does not
1139 * flush disk caches or commits any sort of metadata.
1142 #if defined(DARWIN) || defined(__FreeBSD__)
1145 ret
= ::fdatasync(fd
);
1148 derr
<< __func__
<< " fsync/fdatasync failed: " << cpp_strerror(errno
) << dendl
;
1151 #ifdef HAVE_POSIX_FADVISE
1152 if (cct
->_conf
->filestore_fadvise
)
1153 posix_fadvise(fd
, 0, 0, POSIX_FADV_DONTNEED
);
1157 utime_t lat
= ceph_clock_now() - from
;
1158 dout(20) << "do_write latency " << lat
<< dendl
;
1162 assert(write_pos
== pos
);
1163 assert(write_pos
% header
.alignment
== 0);
1166 Mutex::Locker
locker(finisher_lock
);
1167 journaled_seq
= writing_seq
;
1170 // only if we haven't filled up recently!
1171 if (full_state
!= FULL_NOTFULL
) {
1172 dout(10) << "do_write NOT queueing finisher seq " << journaled_seq
1173 << ", full_commit_seq|full_restart_seq" << dendl
;
1175 if (plug_journal_completions
) {
1176 dout(20) << "do_write NOT queueing finishers through seq " << journaled_seq
1177 << " due to completion plug" << dendl
;
1179 dout(20) << "do_write queueing finishers through seq " << journaled_seq
<< dendl
;
1180 queue_completions_thru(journaled_seq
);
1186 void FileJournal::flush()
1188 dout(10) << "waiting for completions to empty" << dendl
;
1190 Mutex::Locker
l(finisher_lock
);
1191 while (!completions_empty())
1192 finisher_cond
.Wait(finisher_lock
);
1194 dout(10) << "flush waiting for finisher" << dendl
;
1195 finisher
->wait_for_empty();
1196 dout(10) << "flush done" << dendl
;
1200 void FileJournal::write_thread_entry()
1202 dout(10) << "write_thread_entry start" << dendl
;
1205 Mutex::Locker
locker(writeq_lock
);
1206 if (writeq
.empty() && !must_write_header
) {
1209 dout(20) << "write_thread_entry going to sleep" << dendl
;
1210 writeq_cond
.Wait(writeq_lock
);
1211 dout(20) << "write_thread_entry woke up" << dendl
;
1218 Mutex::Locker
locker(aio_lock
);
1219 // should we back off to limit aios in flight? try to do this
1220 // adaptively so that we submit larger aios once we have lots of
1223 // NOTE: our condition here is based on aio_num (protected by
1224 // aio_lock) and throttle_bytes (part of the write queue). when
1225 // we sleep, we *only* wait for aio_num to change, and do not
1226 // wake when more data is queued. this is not strictly correct,
1227 // but should be fine given that we will have plenty of aios in
1228 // flight if we hit this limit to ensure we keep the device
1230 while (aio_num
> 0) {
1231 int exp
= MIN(aio_num
* 2, 24);
1232 long unsigned min_new
= 1ull << exp
;
1233 uint64_t cur
= aio_write_queue_bytes
;
1234 dout(20) << "write_thread_entry aio throttle: aio num " << aio_num
<< " bytes " << aio_bytes
1235 << " ... exp " << exp
<< " min_new " << min_new
1236 << " ... pending " << cur
<< dendl
;
1239 dout(20) << "write_thread_entry deferring until more aios complete: "
1240 << aio_num
<< " aios with " << aio_bytes
<< " bytes needs " << min_new
1241 << " bytes to start a new aio (currently " << cur
<< " pending)" << dendl
;
1242 aio_cond
.Wait(aio_lock
);
1243 dout(20) << "write_thread_entry woke up" << dendl
;
1248 Mutex::Locker
locker(write_lock
);
1249 uint64_t orig_ops
= 0;
1250 uint64_t orig_bytes
= 0;
1253 int r
= prepare_multi_write(bl
, orig_ops
, orig_bytes
);
1254 // Don't care about journal full if stoppping, so drop queue and
1255 // possibly let header get written and loop above to notice stop
1258 dout(20) << "write_thread_entry full and stopping, throw out queue and finish up" << dendl
;
1259 while (!writeq_empty()) {
1260 complete_write(1, peek_write().orig_len
);
1263 print_header(header
);
1266 dout(20) << "write_thread_entry full, going to sleep (waiting for commit)" << dendl
;
1267 commit_cond
.Wait(write_lock
);
1268 dout(20) << "write_thread_entry woke up" << dendl
;
1275 logger
->inc(l_filestore_journal_wr
);
1276 logger
->inc(l_filestore_journal_wr_bytes
, bl
.length());
1287 complete_write(orig_ops
, orig_bytes
);
1290 dout(10) << "write_thread_entry finish" << dendl
;
1294 void FileJournal::do_aio_write(bufferlist
& bl
)
1297 if (cct
->_conf
->journal_write_header_frequency
&&
1298 (((++journaled_since_start
) %
1299 cct
->_conf
->journal_write_header_frequency
) == 0)) {
1300 must_write_header
= true;
1304 if (bl
.length() == 0 && !must_write_header
)
1308 if (must_write_header
) {
1309 must_write_header
= false;
1310 hbp
= prepare_header();
1314 off64_t pos
= write_pos
;
1316 dout(15) << "do_aio_write writing " << pos
<< "~" << bl
.length()
1317 << (hbp
.length() ? " + header":"")
1322 if (pos
+ bl
.length() > header
.max_size
) {
1323 bufferlist first
, second
;
1324 split
= header
.max_size
- pos
;
1325 first
.substr_of(bl
, 0, split
);
1326 second
.substr_of(bl
, split
, bl
.length() - split
);
1327 assert(first
.length() + second
.length() == bl
.length());
1328 dout(10) << "do_aio_write wrapping, first bit at " << pos
<< "~" << first
.length() << dendl
;
1330 if (write_aio_bl(pos
, first
, 0)) {
1331 derr
<< "FileJournal::do_aio_write: write_aio_bl(pos=" << pos
1332 << ") failed" << dendl
;
1335 assert(pos
== header
.max_size
);
1337 // be sneaky: include the header in the second fragment
1338 second
.push_front(hbp
);
1339 pos
= 0; // we included the header
1341 pos
= get_top(); // no header, start after that
1342 if (write_aio_bl(pos
, second
, writing_seq
)) {
1343 derr
<< "FileJournal::do_aio_write: write_aio_bl(pos=" << pos
1344 << ") failed" << dendl
;
1353 if (write_aio_bl(pos
, hbl
, 0)) {
1354 derr
<< "FileJournal::do_aio_write: write_aio_bl(header) failed" << dendl
;
1359 if (write_aio_bl(pos
, bl
, writing_seq
)) {
1360 derr
<< "FileJournal::do_aio_write: write_aio_bl(pos=" << pos
1361 << ") failed" << dendl
;
1367 if (write_pos
== header
.max_size
)
1368 write_pos
= get_top();
1369 assert(write_pos
% header
.alignment
== 0);
1373 * write a buffer using aio
1375 * @param seq seq to trigger when this aio completes. if 0, do not update any state
1378 int FileJournal::write_aio_bl(off64_t
& pos
, bufferlist
& bl
, uint64_t seq
)
1380 dout(20) << "write_aio_bl " << pos
<< "~" << bl
.length() << " seq " << seq
<< dendl
;
1382 while (bl
.length() > 0) {
1383 int max
= MIN(bl
.get_num_buffers(), IOV_MAX
-1);
1384 iovec
*iov
= new iovec
[max
];
1387 for (std::list
<buffer::ptr
>::const_iterator p
= bl
.buffers().begin();
1390 assert(p
!= bl
.buffers().end());
1391 iov
[n
].iov_base
= (void *)p
->c_str();
1392 iov
[n
].iov_len
= p
->length();
1397 bl
.splice(0, len
, &tbl
); // move bytes from bl -> tbl
1399 // lock only aio_queue, current aio, aio_num, aio_bytes, which may be
1400 // modified in check_aio_completion
1402 aio_queue
.push_back(aio_info(tbl
, pos
, bl
.length() > 0 ? 0 : seq
));
1403 aio_info
& aio
= aio_queue
.back();
1406 io_prep_pwritev(&aio
.iocb
, fd
, aio
.iov
, n
, pos
);
1408 dout(20) << "write_aio_bl .. " << aio
.off
<< "~" << aio
.len
1409 << " in " << n
<< dendl
;
1412 aio_bytes
+= aio
.len
;
1414 // need to save current aio len to update write_pos later because current
1415 // aio could be ereased from aio_queue once it is done
1416 uint64_t cur_len
= aio
.len
;
1417 // unlock aio_lock because following io_submit might take time to return
1420 iocb
*piocb
= &aio
.iocb
;
1422 // 2^16 * 125us = ~8 seconds, so max sleep is ~16 seconds
1426 int r
= io_submit(aio_ctx
, 1, &piocb
);
1427 dout(20) << "write_aio_bl io_submit return value: " << r
<< dendl
;
1429 derr
<< "io_submit to " << aio
.off
<< "~" << cur_len
1430 << " got " << cpp_strerror(r
) << dendl
;
1431 if (r
== -EAGAIN
&& attempts
-- > 0) {
1436 check_align(pos
, tbl
);
1437 assert(0 == "io_submit got unexpected error");
1445 write_finish_cond
.Signal();
1451 void FileJournal::write_finish_thread_entry()
1454 dout(10) << "write_finish_thread_entry enter" << dendl
;
1457 Mutex::Locker
locker(aio_lock
);
1458 if (aio_queue
.empty()) {
1461 dout(20) << "write_finish_thread_entry sleeping" << dendl
;
1462 write_finish_cond
.Wait(aio_lock
);
1467 dout(20) << "write_finish_thread_entry waiting for aio(s)" << dendl
;
1469 int r
= io_getevents(aio_ctx
, 1, 16, event
, NULL
);
1472 dout(0) << "io_getevents got " << cpp_strerror(r
) << dendl
;
1475 derr
<< "io_getevents got " << cpp_strerror(r
) << dendl
;
1476 assert(0 == "got unexpected error from io_getevents");
1480 Mutex::Locker
locker(aio_lock
);
1481 for (int i
=0; i
<r
; i
++) {
1482 aio_info
*ai
= (aio_info
*)event
[i
].obj
;
1483 if (event
[i
].res
!= ai
->len
) {
1484 derr
<< "aio to " << ai
->off
<< "~" << ai
->len
1485 << " returned: " << (int)event
[i
].res
<< dendl
;
1486 assert(0 == "unexpected aio error");
1488 dout(10) << "write_finish_thread_entry aio " << ai
->off
1489 << "~" << ai
->len
<< " done" << dendl
;
1492 check_aio_completion();
1495 dout(10) << "write_finish_thread_entry exit" << dendl
;
1501 * check aio_wait for completed aio, and update state appropriately.
1503 void FileJournal::check_aio_completion()
1505 assert(aio_lock
.is_locked());
1506 dout(20) << "check_aio_completion" << dendl
;
1508 bool completed_something
= false, signal
= false;
1509 uint64_t new_journaled_seq
= 0;
1511 list
<aio_info
>::iterator p
= aio_queue
.begin();
1512 while (p
!= aio_queue
.end() && p
->done
) {
1513 dout(20) << "check_aio_completion completed seq " << p
->seq
<< " "
1514 << p
->off
<< "~" << p
->len
<< dendl
;
1516 new_journaled_seq
= p
->seq
;
1517 completed_something
= true;
1520 aio_bytes
-= p
->len
;
1521 aio_queue
.erase(p
++);
1525 if (completed_something
) {
1527 // only if we haven't filled up recently!
1528 Mutex::Locker
locker(finisher_lock
);
1529 journaled_seq
= new_journaled_seq
;
1530 if (full_state
!= FULL_NOTFULL
) {
1531 dout(10) << "check_aio_completion NOT queueing finisher seq " << journaled_seq
1532 << ", full_commit_seq|full_restart_seq" << dendl
;
1534 if (plug_journal_completions
) {
1535 dout(20) << "check_aio_completion NOT queueing finishers through seq " << journaled_seq
1536 << " due to completion plug" << dendl
;
1538 dout(20) << "check_aio_completion queueing finishers through seq " << journaled_seq
<< dendl
;
1539 queue_completions_thru(journaled_seq
);
1544 // maybe write queue was waiting for aio count to drop?
1550 int FileJournal::prepare_entry(vector
<ObjectStore::Transaction
>& tls
, bufferlist
* tbl
) {
1551 dout(10) << "prepare_entry " << tls
<< dendl
;
1552 int data_len
= cct
->_conf
->journal_align_min_size
- 1;
1553 int data_align
= -1; // -1 indicates that we don't care about the alignment
1555 for (vector
<ObjectStore::Transaction
>::iterator p
= tls
.begin();
1556 p
!= tls
.end(); ++p
) {
1557 if ((int)(*p
).get_data_length() > data_len
) {
1558 data_len
= (*p
).get_data_length();
1559 data_align
= ((*p
).get_data_alignment() - bl
.length()) & ~CEPH_PAGE_MASK
;
1563 if (tbl
->length()) {
1564 bl
.claim_append(*tbl
);
1566 // add it this entry
1568 unsigned head_size
= sizeof(entry_header_t
);
1569 off64_t base_size
= 2*head_size
+ bl
.length();
1570 memset(&h
, 0, sizeof(h
));
1571 if (data_align
>= 0)
1572 h
.pre_pad
= ((unsigned int)data_align
- (unsigned int)head_size
) & ~CEPH_PAGE_MASK
;
1573 off64_t size
= ROUND_UP_TO(base_size
+ h
.pre_pad
, header
.alignment
);
1574 unsigned post_pad
= size
- base_size
- h
.pre_pad
;
1575 h
.len
= bl
.length();
1576 h
.post_pad
= post_pad
;
1577 h
.crc32c
= bl
.crc32c(0);
1578 dout(10) << " len " << bl
.length() << " -> " << size
1579 << " (head " << head_size
<< " pre_pad " << h
.pre_pad
1580 << " bl " << bl
.length() << " post_pad " << post_pad
<< " tail " << head_size
<< ")"
1581 << " (bl alignment " << data_align
<< ")"
1585 ebl
.append((const char*)&h
, sizeof(h
));
1587 ebl
.push_back(buffer::create_static(h
.pre_pad
, zero_buf
));
1590 ebl
.claim_append(bl
, buffer::list::CLAIM_ALLOW_NONSHAREABLE
); // potential zero-copy
1592 ebl
.push_back(buffer::create_static(h
.post_pad
, zero_buf
));
1595 ebl
.append((const char*)&h
, sizeof(h
));
1597 ebl
.rebuild_aligned(CEPH_DIRECTIO_ALIGNMENT
);
1602 void FileJournal::submit_entry(uint64_t seq
, bufferlist
& e
, uint32_t orig_len
,
1603 Context
*oncommit
, TrackedOpRef osd_op
)
1606 dout(5) << "submit_entry seq " << seq
1607 << " len " << e
.length()
1608 << " (" << oncommit
<< ")" << dendl
;
1609 assert(e
.length() > 0);
1610 assert(e
.length() < header
.max_size
);
1613 osd_op
->mark_event("commit_queued_for_journal_write");
1615 logger
->inc(l_filestore_journal_queue_bytes
, orig_len
);
1616 logger
->inc(l_filestore_journal_queue_ops
, 1);
1619 throttle
.register_throttle_seq(seq
, e
.length());
1621 logger
->inc(l_filestore_journal_ops
, 1);
1622 logger
->inc(l_filestore_journal_bytes
, e
.length());
1626 osd_op
->mark_event("commit_queued_for_journal_write");
1627 if (osd_op
->store_trace
) {
1628 osd_op
->journal_trace
.init("journal", &trace_endpoint
, &osd_op
->store_trace
);
1629 osd_op
->journal_trace
.event("submit_entry");
1630 osd_op
->journal_trace
.keyval("seq", seq
);
1634 Mutex::Locker
l1(writeq_lock
);
1636 Mutex::Locker
l2(aio_lock
);
1638 Mutex::Locker
l3(completions_lock
);
1641 aio_write_queue_ops
++;
1642 aio_write_queue_bytes
+= e
.length();
1646 completions
.push_back(
1648 seq
, oncommit
, ceph_clock_now(), osd_op
));
1650 writeq_cond
.Signal();
1651 writeq
.push_back(write_item(seq
, e
, orig_len
, osd_op
));
1653 osd_op
->journal_trace
.keyval("queue depth", writeq
.size());
1657 bool FileJournal::writeq_empty()
1659 Mutex::Locker
locker(writeq_lock
);
1660 return writeq
.empty();
1663 FileJournal::write_item
&FileJournal::peek_write()
1665 assert(write_lock
.is_locked());
1666 Mutex::Locker
locker(writeq_lock
);
1667 return writeq
.front();
1670 void FileJournal::pop_write()
1672 assert(write_lock
.is_locked());
1673 Mutex::Locker
locker(writeq_lock
);
1675 logger
->dec(l_filestore_journal_queue_bytes
, writeq
.front().orig_len
);
1676 logger
->dec(l_filestore_journal_queue_ops
, 1);
1681 void FileJournal::batch_pop_write(list
<write_item
> &items
)
1683 assert(write_lock
.is_locked());
1685 Mutex::Locker
locker(writeq_lock
);
1688 for (auto &&i
: items
) {
1690 logger
->dec(l_filestore_journal_queue_bytes
, i
.orig_len
);
1691 logger
->dec(l_filestore_journal_queue_ops
, 1);
1696 void FileJournal::batch_unpop_write(list
<write_item
> &items
)
1698 assert(write_lock
.is_locked());
1699 for (auto &&i
: items
) {
1701 logger
->inc(l_filestore_journal_queue_bytes
, i
.orig_len
);
1702 logger
->inc(l_filestore_journal_queue_ops
, 1);
1705 Mutex::Locker
locker(writeq_lock
);
1706 writeq
.splice(writeq
.begin(), items
);
1709 void FileJournal::commit_start(uint64_t seq
)
1711 dout(10) << "commit_start" << dendl
;
1714 switch (full_state
) {
1719 if (seq
>= journaled_seq
) {
1720 dout(1) << " FULL_FULL -> FULL_WAIT. commit_start on seq "
1721 << seq
<< " > journaled_seq " << journaled_seq
1722 << ", moving to FULL_WAIT."
1724 full_state
= FULL_WAIT
;
1726 dout(1) << "FULL_FULL commit_start on seq "
1727 << seq
<< " < journaled_seq " << journaled_seq
1728 << ", remaining in FULL_FULL"
1734 dout(1) << " FULL_WAIT -> FULL_NOTFULL. journal now active, setting completion plug." << dendl
;
1735 full_state
= FULL_NOTFULL
;
1736 plug_journal_completions
= true;
1742 *send discard command to joural block deivce
1744 void FileJournal::do_discard(int64_t offset
, int64_t end
)
1746 dout(10) << __func__
<< "trim(" << offset
<< ", " << end
<< dendl
;
1748 offset
= ROUND_UP_TO(offset
, block_size
);
1751 end
= ROUND_UP_TO(end
- block_size
, block_size
);
1752 assert(end
>= offset
);
1754 if (block_device_discard(fd
, offset
, end
- offset
) < 0)
1755 dout(1) << __func__
<< "ioctl(BLKDISCARD) error:" << cpp_strerror(errno
) << dendl
;
1758 void FileJournal::committed_thru(uint64_t seq
)
1760 Mutex::Locker
locker(write_lock
);
1762 auto released
= throttle
.flush(seq
);
1764 logger
->dec(l_filestore_journal_ops
, released
.first
);
1765 logger
->dec(l_filestore_journal_bytes
, released
.second
);
1768 if (seq
< last_committed_seq
) {
1769 dout(5) << "committed_thru " << seq
<< " < last_committed_seq " << last_committed_seq
<< dendl
;
1770 assert(seq
>= last_committed_seq
);
1773 if (seq
== last_committed_seq
) {
1774 dout(5) << "committed_thru " << seq
<< " == last_committed_seq " << last_committed_seq
<< dendl
;
1778 dout(5) << "committed_thru " << seq
<< " (last_committed_seq " << last_committed_seq
<< ")" << dendl
;
1779 last_committed_seq
= seq
;
1783 Mutex::Locker
locker(finisher_lock
);
1784 queue_completions_thru(seq
);
1785 if (plug_journal_completions
&& seq
>= header
.start_seq
) {
1786 dout(10) << " removing completion plug, queuing completions thru journaled_seq " << journaled_seq
<< dendl
;
1787 plug_journal_completions
= false;
1788 queue_completions_thru(journaled_seq
);
1792 // adjust start pointer
1793 while (!journalq
.empty() && journalq
.front().first
<= seq
) {
1794 journalq
.pop_front();
1797 int64_t old_start
= header
.start
;
1798 if (!journalq
.empty()) {
1799 header
.start
= journalq
.front().second
;
1800 header
.start_seq
= journalq
.front().first
;
1802 header
.start
= write_pos
;
1803 header
.start_seq
= seq
+ 1;
1807 dout(10) << __func__
<< " will trim (" << old_start
<< ", " << header
.start
<< ")" << dendl
;
1808 if (old_start
< header
.start
)
1809 do_discard(old_start
, header
.start
- 1);
1811 do_discard(old_start
, header
.max_size
- 1);
1812 do_discard(get_top(), header
.start
- 1);
1816 must_write_header
= true;
1817 print_header(header
);
1819 // committed but unjournaled items
1820 while (!writeq_empty() && peek_write().seq
<= seq
) {
1821 dout(15) << " dropping committed but unwritten seq " << peek_write().seq
1822 << " len " << peek_write().bl
.length()
1824 complete_write(1, peek_write().orig_len
);
1828 commit_cond
.Signal();
1830 dout(10) << "committed_thru done" << dendl
;
1834 void FileJournal::complete_write(uint64_t ops
, uint64_t bytes
)
1836 dout(5) << __func__
<< " finished " << ops
<< " ops and "
1837 << bytes
<< " bytes" << dendl
;
1840 int FileJournal::make_writeable()
1842 dout(10) << __func__
<< dendl
;
1843 int r
= set_throttle_params();
1852 write_pos
= read_pos
;
1854 write_pos
= get_top();
1857 must_write_header
= true;
1863 int FileJournal::set_throttle_params()
1866 bool valid
= throttle
.set_params(
1867 cct
->_conf
->journal_throttle_low_threshhold
,
1868 cct
->_conf
->journal_throttle_high_threshhold
,
1869 cct
->_conf
->filestore_expected_throughput_bytes
,
1870 cct
->_conf
->journal_throttle_high_multiple
,
1871 cct
->_conf
->journal_throttle_max_multiple
,
1872 header
.max_size
- get_top(),
1876 derr
<< "tried to set invalid params: "
1880 return valid
? 0 : -EINVAL
;
1883 const char** FileJournal::get_tracked_conf_keys() const
1885 static const char *KEYS
[] = {
1886 "journal_throttle_low_threshhold",
1887 "journal_throttle_high_threshhold",
1888 "journal_throttle_high_multiple",
1889 "journal_throttle_max_multiple",
1890 "filestore_expected_throughput_bytes",
1895 void FileJournal::wrap_read_bl(
1903 while (pos
>= header
.max_size
)
1904 pos
= pos
+ get_top() - header
.max_size
;
1907 if (pos
+ olen
> header
.max_size
)
1908 len
= header
.max_size
- pos
; // partial
1912 int64_t actual
= ::lseek64(fd
, pos
, SEEK_SET
);
1913 assert(actual
== pos
);
1915 bufferptr bp
= buffer::create(len
);
1916 int r
= safe_read_exact(fd
, bp
.c_str(), len
);
1918 derr
<< "FileJournal::wrap_read_bl: safe_read_exact " << pos
<< "~" << len
<< " returned "
1922 bl
->push_back(std::move(bp
));
1926 if (pos
>= header
.max_size
)
1927 pos
= pos
+ get_top() - header
.max_size
;
1932 bool FileJournal::read_entry(
1939 uint64_t seq
= next_seq
;
1942 dout(2) << "read_entry -- not readable" << dendl
;
1946 off64_t pos
= read_pos
;
1947 off64_t next_pos
= pos
;
1949 read_entry_result result
= do_read_entry(
1955 if (result
== SUCCESS
) {
1956 journalq
.push_back( pair
<uint64_t,off64_t
>(seq
, pos
));
1957 uint64_t amount_to_take
=
1960 (header
.max_size
- pos
) + (next_pos
- get_top());
1961 throttle
.take(amount_to_take
);
1962 throttle
.register_throttle_seq(next_seq
, amount_to_take
);
1964 logger
->inc(l_filestore_journal_ops
, 1);
1965 logger
->inc(l_filestore_journal_bytes
, amount_to_take
);
1967 if (next_seq
> seq
) {
1970 read_pos
= next_pos
;
1972 if (seq
> journaled_seq
)
1973 journaled_seq
= seq
;
1977 derr
<< "do_read_entry(" << pos
<< "): " << ss
.str() << dendl
;
1980 if (seq
&& seq
< header
.committed_up_to
) {
1981 derr
<< "Unable to read past sequence " << seq
1982 << " but header indicates the journal has committed up through "
1983 << header
.committed_up_to
<< ", journal is corrupt" << dendl
;
1984 if (cct
->_conf
->journal_ignore_corruption
) {
1993 dout(2) << "No further valid entries found, journal is most likely valid"
1998 FileJournal::read_entry_result
FileJournal::do_read_entry(
2004 entry_header_t
*_h
) const
2006 off64_t cur_pos
= init_pos
;
2015 wrap_read_bl(cur_pos
, sizeof(*h
), &hbl
, &_next_pos
);
2016 h
= reinterpret_cast<entry_header_t
*>(hbl
.c_str());
2018 if (!h
->check_magic(cur_pos
, header
.get_fsid64())) {
2019 dout(25) << "read_entry " << init_pos
2020 << " : bad header magic, end of journal" << dendl
;
2022 *ss
<< "bad header magic";
2024 *next_pos
= init_pos
+ (4<<10); // check 4k ahead
2025 return MAYBE_CORRUPT
;
2027 cur_pos
= _next_pos
;
2031 cur_pos
+= h
->pre_pad
;
2034 wrap_read_bl(cur_pos
, h
->len
, bl
, &cur_pos
);
2037 cur_pos
+= h
->post_pad
;
2042 wrap_read_bl(cur_pos
, sizeof(*f
), &fbl
, &cur_pos
);
2043 f
= reinterpret_cast<entry_header_t
*>(fbl
.c_str());
2044 if (memcmp(f
, h
, sizeof(*f
))) {
2046 *ss
<< "bad footer magic, partial entry";
2048 *next_pos
= cur_pos
;
2049 return MAYBE_CORRUPT
;
2052 if ((header
.flags
& header_t::FLAG_CRC
) || // if explicitly enabled (new journal)
2053 h
->crc32c
!= 0) { // newer entry in old journal
2054 uint32_t actual_crc
= bl
->crc32c(0);
2055 if (actual_crc
!= h
->crc32c
) {
2057 *ss
<< "header crc (" << h
->crc32c
2058 << ") doesn't match body crc (" << actual_crc
<< ")";
2060 *next_pos
= cur_pos
;
2061 return MAYBE_CORRUPT
;
2066 dout(2) << "read_entry " << init_pos
<< " : seq " << h
->seq
2067 << " " << h
->len
<< " bytes"
2076 *next_pos
= cur_pos
;
2081 assert(cur_pos
% header
.alignment
== 0);
2085 void FileJournal::reserve_throttle_and_backoff(uint64_t count
)
2087 throttle
.get(count
);
2090 void FileJournal::get_header(
2091 uint64_t wanted_seq
,
2095 off64_t pos
= header
.start
;
2096 off64_t next_pos
= pos
;
2099 dout(2) << __func__
<< dendl
;
2103 read_entry_result result
= do_read_entry(
2110 if (result
== FAILURE
|| result
== MAYBE_CORRUPT
)
2112 if (seq
== wanted_seq
) {
2118 ceph_abort(); // not reachable
2121 void FileJournal::corrupt(
2125 dout(2) << __func__
<< dendl
;
2126 if (corrupt_at
>= header
.max_size
)
2127 corrupt_at
= corrupt_at
+ get_top() - header
.max_size
;
2129 int64_t actual
= ::lseek64(fd
, corrupt_at
, SEEK_SET
);
2130 assert(actual
== corrupt_at
);
2133 int r
= safe_read_exact(fd
, buf
, 1);
2136 actual
= ::lseek64(wfd
, corrupt_at
, SEEK_SET
);
2137 assert(actual
== corrupt_at
);
2140 r
= safe_write(wfd
, buf
, 1);
2144 void FileJournal::corrupt_payload(
2148 dout(2) << __func__
<< dendl
;
2151 get_header(seq
, &pos
, &h
);
2152 off64_t corrupt_at
=
2153 pos
+ sizeof(entry_header_t
) + h
.pre_pad
;
2154 corrupt(wfd
, corrupt_at
);
2158 void FileJournal::corrupt_footer_magic(
2162 dout(2) << __func__
<< dendl
;
2165 get_header(seq
, &pos
, &h
);
2166 off64_t corrupt_at
=
2167 pos
+ sizeof(entry_header_t
) + h
.pre_pad
+
2168 h
.len
+ h
.post_pad
+
2169 (reinterpret_cast<char*>(&h
.magic2
) - reinterpret_cast<char*>(&h
));
2170 corrupt(wfd
, corrupt_at
);
2174 void FileJournal::corrupt_header_magic(
2178 dout(2) << __func__
<< dendl
;
2181 get_header(seq
, &pos
, &h
);
2182 off64_t corrupt_at
=
2184 (reinterpret_cast<char*>(&h
.magic2
) - reinterpret_cast<char*>(&h
));
2185 corrupt(wfd
, corrupt_at
);
2188 off64_t
FileJournal::get_journal_size_estimate()
2190 off64_t size
, start
= header
.start
;
2191 if (write_pos
< start
) {
2192 size
= (max_size
- start
) + write_pos
;
2194 size
= write_pos
- start
;
2196 dout(20) << __func__
<< " journal size=" << size
<< dendl
;