Обсуждение: [bug fix] pg_rewind takes long time because it mistakenly copiesdata files

Hello,

Our customer reported that pg_rewind took many hours to synchronize 400GB of data, even if the new primary doesn't
performany updates.  The attached patch fixes that.
 

The cause was that pg_rewind failed to recognize data files in tablespace directories, resulting in the full copy of
thosefiles instead of WAL replay.
 

Regards
Takayuki Tsunakawa

On Mon, Feb 26, 2018 at 06:01:43AM +0000, Tsunakawa, Takayuki wrote:
> The cause was that pg_rewind failed to recognize data files in
> tablespace directories, resulting in the full copy of those files
> instead of WAL replay.

Ouch.  Confirmed.

If I test pg_rewind with a tablespace (primary and standby on same
server, base backup taken using --tablespace-mapping), then I bump
immediately into an assertion failure:
(lldb) bt
* thread #1: tid = 0x0000, 0x00007fff99aa3f06 libsystem_kernel.dylib`__pthread_kill + 10, stop reason = signal SIGSTOP
  * frame #0: 0x00007fff99aa3f06 libsystem_kernel.dylib`__pthread_kill + 10
    frame #1: 0x00007fff8a30e4ec libsystem_pthread.dylib`pthread_kill + 90
    frame #2: 0x00007fff9a8d46df libsystem_c.dylib`abort + 129
    frame #3: 0x00007fff9a89bdd8 libsystem_c.dylib`__assert_rtn + 321
    frame #4: 0x000000010f7680f6 pg_rewind`process_block_change(forknum=MAIN_FORKNUM, rnode=(spcNode = 16385, dbNode =
16384,relNode = 16386), blkno=4424) + 406 at filemap.c:374 
    frame #5: 0x000000010f760c9f pg_rewind`extractPageInfo(record=0x00007fc742800600) + 399 at parsexlog.c:401
    frame #6: 0x000000010f76055e pg_rewind`extractPageMap(datadir="/Users/XXX/data/5432", startpoint=0, tliIndex=0,
endpoint=134217768)+ 270 at parsexlog.c:97 
    frame #7: 0x000000010f75f48c pg_rewind`main(argc=5, argv=0x00007fff504a1450) + 1964 at pg_rewind.c:300
    frame #8: 0x00007fff88daf5ad libdyld.dylib`start + 1
(lldb) up 4
frame #4: 0x000000010f7680f6 pg_rewind`process_block_change(forknum=MAIN_FORKNUM, rnode=(spcNode = 16385, dbNode =
16384,relNode = 16386), blkno=4424) + 406 at filemap.c:374 
   371
   372         if (entry)
   373         {
-> 374             Assert(entry->isrelfile);
   375
   376             switch (entry->action)
   377             {
(lldb) p *entry
(file_entry_t) $0 = {
  path = 0x00007fc742424400 "pg_tblspc/16385/PG_11_201802061/16384/16386"
  type = FILE_TYPE_REGULAR
  action = FILE_ACTION_COPY
  oldsize = 0
  newsize = 36249600
  isrelfile = '\0'
  pagemap = (bitmap = <no value available>, bitmapsize = 0)
  link_target = 0x00007fc742424430 <no value available>
  next = 0x0000000000000000
}

Your patch is able to fix that.  I have also checked that after
diverging the promoted server with more data and inserting data on the
old primary then the correct set of blocks from the tablespace is fetched as
well by pg_rewind.  This patch also behaves correctly when creating a
new relation on the promoted server as it copies the whole relation. In
short your patch looks good to me.

Creating a test case for this patch would be nice, however this needs a
bit of work so as the tablespace map can be used with pg_basebackup and
PostgresNode.pm (or use raw pg_basebackup commands in pg_rewind tests):
- PostgresNode::init_from_backup needs to be able to understand extra
options given by caller for pg_basebackup.
- RewindTest::create_standby should be extended with extra arguments
given for previous extension.
:(
--
Michael

From: Michael Paquier [mailto:michael@paquier.xyz]
> Your patch is able to fix that.  I have also checked that after diverging
> the promoted server with more data and inserting data on the old primary
> then the correct set of blocks from the tablespace is fetched as well by
> pg_rewind.  This patch also behaves correctly when creating a new relation
> on the promoted server as it copies the whole relation. In short your patch
> looks good to me.

How quick, I was surprised.  Thank you so much!  I'd be glad if you could be the reviewer in CF:

https://commitfest.postgresql.org/17/1542/


> Creating a test case for this patch would be nice, however this needs a
> bit of work so as the tablespace map can be used with pg_basebackup and
> PostgresNode.pm (or use raw pg_basebackup commands in pg_rewind tests):
> - PostgresNode::init_from_backup needs to be able to understand extra
> options given by caller for pg_basebackup.
> - RewindTest::create_standby should be extended with extra arguments given
> for previous extension.
> :(

That sounds difficult from your way of saying... but this may be a good opportunity to look into the TAP tests.

Regards
Takayuki Tsunakawa

On Mon, Feb 26, 2018 at 08:13:02AM +0000, Tsunakawa, Takayuki wrote:
> From: Michael Paquier [mailto:michael@paquier.xyz]
>> Your patch is able to fix that.  I have also checked that after diverging
>> the promoted server with more data and inserting data on the old primary
>> then the correct set of blocks from the tablespace is fetched as well by
>> pg_rewind.  This patch also behaves correctly when creating a new relation
>> on the promoted server as it copies the whole relation. In short your patch
>> looks good to me.
>
> How quick, I was surprised.  Thank you so much!  I'd be glad if you could be the reviewer in CF:
>
> https://commitfest.postgresql.org/17/1542/

Sure.

>> Creating a test case for this patch would be nice, however this needs a
>> bit of work so as the tablespace map can be used with pg_basebackup and
>> PostgresNode.pm (or use raw pg_basebackup commands in pg_rewind tests):
>> - PostgresNode::init_from_backup needs to be able to understand extra
>> options given by caller for pg_basebackup.
>> - RewindTest::create_standby should be extended with extra arguments given
>> for previous extension.
>> :(
>
> That sounds difficult from your way of saying... but this may be a
> good opportunity to look into the TAP tests.

Anything like that would be work only for HEAD I think as that's a bit
of refactoring.  And indeed it could give you a good introduction to the
TAP facility.
--
Michael

On Mon, Feb 26, 2018 at 06:24:02PM +0900, Michael Paquier wrote:
> Anything like that would be work only for HEAD I think as that's a bit
> of refactoring.  And indeed it could give you a good introduction to the
> TAP facility.

So I would propose to just do that later.  I have looked a second time
at your patch, attached is the set of tests I have run:
- With assertions enabled, I see pg_rewind failing on an assertion as
mentioned upthread.
- With assertion disabled (look at rewind_logs.log as generated by the
script attached), then all the files from the tablespace are copied with
all the file chunks fetched at the second phase of the rewind.
- With the patch, both the assertion and the handling of tablespace
files are fixed.  Again, in the logs produced by the script you would
see that tablespace files are not completely copied anymore, and that
only portions of them are.

I have one small comment though.  The comment block at the beginning of
isRelDataFile() refers to "pg_tblspc/<tblspc oid>/PG_9.4_201403261/".
Could you switch that to "pg_tblspc/<tblspc oid>/<tblspc version>"?
That's not directly the fault of your patch, but as long as we are on
it..

So I am marking this as ready for committer.  Thanks for the report,
Tsunakawa-san!
--
Michael

From: Michael Paquier [mailto:michael@paquier.xyz]
> So I would propose to just do that later.  I have looked a second time at
> your patch, attached is the set of tests I have run:

Thanks so much, that has helped me a lot!

> I have one small comment though.  The comment block at the beginning of
> isRelDataFile() refers to "pg_tblspc/<tblspc oid>/PG_9.4_201403261/".
> Could you switch that to "pg_tblspc/<tblspc oid>/<tblspc version>"?
> That's not directly the fault of your patch, but as long as we are on it..

Done, thanks again for marking the CF entry.

Regards
Takayuki Tsunakawa

On Wed, Feb 28, 2018 at 3:58 PM, Tsunakawa, Takayuki
<tsunakawa.takay@jp.fujitsu.com> wrote:
> From: Michael Paquier [mailto:michael@paquier.xyz]
>> So I would propose to just do that later.  I have looked a second time at
>> your patch, attached is the set of tests I have run:
>
> Thanks so much, that has helped me a lot!
>
>> I have one small comment though.  The comment block at the beginning of
>> isRelDataFile() refers to "pg_tblspc/<tblspc oid>/PG_9.4_201403261/".
>> Could you switch that to "pg_tblspc/<tblspc oid>/<tblspc version>"?
>> That's not directly the fault of your patch, but as long as we are on it..
>
> Done, thanks again for marking the CF entry.

Thanks for the patch! Pushed.

Regards,

-- 
Fujii Masao

From: Fujii Masao [mailto:masao.fujii@gmail.com]
> Thanks for the patch! Pushed.

Thank you.  I'm glad to see you again on this list.

Regards
Takayuki Tsunakawa