Re: [HACKERS] An attempt to reduce WALWriteLock contention
On Thu, Jun 22, 2017 at 6:54 AM, Andres Freund wrote: > On 2017-06-21 00:57:32 -0700, jasrajd wrote: >> We are also seeing contention on the walwritelock and repeated writes to the >> same offset if we move the flush outside the lock in the Azure environment. >> pgbench doesn't scale beyond ~8 cores without saturating the IOPs or >> bandwidth. Is there more work being done in this area? > That should not happen if the writes from various backends are combined in some way. However, it is not very clear what exactly you have done as part of taking flush calls out of walwritelock. Can you share patch or some details about how you have done it and how have you measured the contention you are seeing? -- With Regards, Amit Kapila. EnterpriseDB: http://www.enterprisedb.com -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On 2017-06-22 04:16, Michael Paquier wrote: On Wed, Jun 21, 2017 at 4:57 PM, jasrajd wrote: We are also seeing contention on the walwritelock and repeated writes to the same offset if we move the flush outside the lock in the Azure environment. pgbench doesn't scale beyond ~8 cores without saturating the IOPs or bandwidth. Is there more work being done in this area? As of now, there is no patch in the development queue for Postgres 11 that is dedicated to this particularly lock contention. There is a patch for LWlocks in general with power PC, but that's all: https://commitfest.postgresql.org/14/984/ Not sure if Kuntal has plans to submit again this patch. It is actually a bit sad to not see things moving on and use an approach to group flushes. -- Michael There is also patch against LWLock degradation on NUMA : https://commitfest.postgresql.org/14/1166/ But they are both about LWLock itself, and not its usage. -- Sokolov Yura Postgres Professional: https://postgrespro.ru The Russian Postgres Company -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On Thu, Jun 22, 2017 at 6:46 AM, Michael Paquier wrote: > On Wed, Jun 21, 2017 at 4:57 PM, jasrajd wrote: >> We are also seeing contention on the walwritelock and repeated writes to the >> same offset if we move the flush outside the lock in the Azure environment. >> pgbench doesn't scale beyond ~8 cores without saturating the IOPs or >> bandwidth. Is there more work being done in this area? > > As of now, there is no patch in the development queue for Postgres 11 > that is dedicated to this particularly lock contention. There is a > patch for LWlocks in general with power PC, but that's all: > https://commitfest.postgresql.org/14/984/ > > Not sure if Kuntal has plans to submit again this patch. It is > actually a bit sad to not see things moving on and use an approach to > group flushes. As of now, I've no plans to re-submit the patch. Actually, I'm not sure what I should try next. I would love to get some advice/direction regarding this. -- Thanks & Regards, Kuntal Ghosh EnterpriseDB: http://www.enterprisedb.com -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On Thu, Dec 22, 2016 at 11:29 PM, Tomas Vondra wrote: > > How do these counts compare to the other wait events? For example > CLogControlLock, which is what Amit's patch [1] is about? > > [1] > https://www.postgresql.org/message-id/flat/84c22fbb-b9c4-a02f-384b-b4feb2c67193%402ndquadrant.com > Hello Tomas, I'm really sorry for this late reply. I've somehow missed the thread. Actually, I've seen some performance improvement with the CLogControlLock patch. But, then it turned out all the improvements were because of the CLogControlLock patch alone. -- Thanks & Regards, Kuntal Ghosh EnterpriseDB: http://www.enterprisedb.com -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On 2017-06-21 00:57:32 -0700, jasrajd wrote: > We are also seeing contention on the walwritelock and repeated writes to the > same offset if we move the flush outside the lock in the Azure environment. > pgbench doesn't scale beyond ~8 cores without saturating the IOPs or > bandwidth. Is there more work being done in this area? I kind of doubt that scalability limit is directly related to this patch - I've seen postgres scale furhter without that lock becoming the prime issue. What exactly are you measuring / observing? Andres Freund -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On Wed, Jun 21, 2017 at 4:57 PM, jasrajd wrote: > We are also seeing contention on the walwritelock and repeated writes to the > same offset if we move the flush outside the lock in the Azure environment. > pgbench doesn't scale beyond ~8 cores without saturating the IOPs or > bandwidth. Is there more work being done in this area? As of now, there is no patch in the development queue for Postgres 11 that is dedicated to this particularly lock contention. There is a patch for LWlocks in general with power PC, but that's all: https://commitfest.postgresql.org/14/984/ Not sure if Kuntal has plans to submit again this patch. It is actually a bit sad to not see things moving on and use an approach to group flushes. -- Michael -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
We are also seeing contention on the walwritelock and repeated writes to the same offset if we move the flush outside the lock in the Azure environment. pgbench doesn't scale beyond ~8 cores without saturating the IOPs or bandwidth. Is there more work being done in this area? -- View this message in context: http://www.postgresql-archive.org/An-attempt-to-reduce-WALWriteLock-contention-tp5935907p5967786.html Sent from the PostgreSQL - hackers mailing list archive at Nabble.com. -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] An attempt to reduce WALWriteLock contention
On 12/22/2016 04:00 PM, Kuntal Ghosh wrote: Hello all, > ... > \t select wait_event_type, wait_event from pg_stat_activity where pid != pg_backend_pid(); \watch 0.5 HEAD 48642 LWLockNamed | WALWriteLock With Patch -- 31889 LWLockNamed | WALFlushLock 25212 LWLockNamed | WALWriteLock How do these counts compare to the other wait events? For example CLogControlLock, which is what Amit's patch [1] is about? [1] https://www.postgresql.org/message-id/flat/84c22fbb-b9c4-a02f-384b-b4feb2c67193%402ndquadrant.com regards -- Tomas Vondra http://www.2ndQuadrant.com PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
[HACKERS] An attempt to reduce WALWriteLock contention
Hello all, In a recent post[1] by Robert, wait events for different LWLOCKS have been analyzed. The results clearly indicate a significant lock contention overhead on WAL Write locks. To get an idea of this overhead, we did the following two tests. 1. Hacked the code to comment out WAL write and flush calls to see the overhead of WAL writing. The TPS for read-write pgbench tests at 300 scale factor with 64 client count increased from 27871 to 45068. 2. Hacked the code to comment out WAL flush calls to see the overhead of WAL flushing (fsync=off). The TPS for read-write pgbench tests at 300 scale factor with 64 client count increased from 27871 to 41835. All the tests have been performed for 15 minutes with following pg configurations: max_wal_size: 40GB checkpoint_timeout: 15mins maintenance_work_mem: 4GB checkpoint_completion_target: 0.9 Shared buffer: 8GB (Other settings have default values) >From above experiments, it is clear that flush is the main cost in WAL writing which is no surprise, but still, the above data shows the exact overhead of flush. Robert and Amit suggested (in offline discussions) using separate WALFlushLock to flush the WAL data. The idea is to take WAL flush calls out of WAL Write Lock and introduce a new lock (WAL Flush Lock) to flush the data. This should allow simultaneous os writes when a fsync is in progress. LWLockAcquireOrWait is used for the newly introduced WAL Flush Lock to accumulate flush calls. We did a pgbench read/write (s.f. 300) test with above configurations for various clients. But, we didn't see any performance improvements, rather it decreased by 10%-12%. Hence to measure the wait events, we performed a run for 30 minutes with 64 clients. \t select wait_event_type, wait_event from pg_stat_activity where pid != pg_backend_pid(); \watch 0.5 HEAD 48642 LWLockNamed | WALWriteLock With Patch -- 31889 LWLockNamed | WALFlushLock 25212 LWLockNamed | WALWriteLock The contention on WAL Write Lock was reduced, but together with WAL Flush lock, the total contention got increased. We also measured the number of times fsync() and write() have been called for a 10-minutes pgbench read/write test with 16 clients. We noticed a huge increase in write() system calls and this is happening as we've reduced the contention on WAL Write Lock. Due to reduced contention on WAL Write Lock, lot of backends are going for small os writes, sometimes on same 8KB page, i.e., write calls are not properly accumulated. For example, backend 1 - 1 KB write() - 15-20 micro secs backend 2 - 1 KB write() - 15-20 micro secs backend 3 - 1 KB write() - 15-20 micro secs backend 4 - 1 KB write() - 15-20 micro secs But, if we accumulate these 4 requests, 4KB can be written in 50-60 micro secs. Apart from that, we are also paying for lock acquire and lock release for os write and lseek(). For the same reason, when a fsync is going, we are not able to accumulate sufficient data for next fsync. This also increases the contention on WAL Flush Lock. So, we tried adding delay(pg_usleep) before flush/write to accumulate data. But, this severely increases the contention on WAL flush locks. To reduce the contention on WAL Write Lock further, Amit suggested the following change on top of the existing patch: Backend as Write Leader: Except one proc, all other proc's will wait for their write location to be written in OS buffer. Each proc will advertise it's write location and wait on the semaphore to check whether it's write location has been completed. Only the leader will compete for WALWriteLock.After data is written, it wakes all the procs for which it has written the WAL and once done with waking it will release the WALWriteLock. Ashutosh and Amit have helped a lot for the implementation of the above idea. Even after this idea, we didn't see any noticeable performance improvement with synchronous_commit=on mode, however there was no regression. Again, to measure the wait events, we performed a 30 minutes run with 64 clients. (pgbench r/w test with s.f. 300) \t select wait_event_type, wait_event from pg_stat_activity where pid != pg_backend_pid(); \watch 0.5 HEAD 48642 LWLockNamed | WALWriteLock With Patch -- 38952 LWLockNamed | WALFlushLock 1679 LWLockNamed | WALWriteLock We reduced the contention on WAL write locks. The reason is that only the group leader is competing for write lock on behalf of a group of procs. Still, the number of small write requests is not reduced. Finally, we performed some tests with synchronous_commit=off and data doesn't fit in shared buffer. This should accumulate the data properly for write without waiting on some locks or semaphores. Besides, write and fsync can be done simultaneously. Next results are for various scale factors and shared buffers. (Please see below for system configuration): ./pgbench -c $threads -j $threads -T 900 -M prepared postgres n