Use pg_test_fsync test which wal_sync_method is fastest in your filesystem
wal_sync_method 提供了多种fsync的方法。各种方法的效率可能存在差异。
#wal_sync_method = fsync
# the default is the first option
# supported by the operating system:
# open_datasync
# fdatasync (default on Linux)
# fsync
# fsync_writethrough
# open_sync
PostgreSQL提供了一个util用来测试系统所有支持的fsync() 方法下的性能 , 包括是否支持multi-processes fsync同一个文件。
原文引用 :
* Test whether fsync can sync data written on a different descriptor for
* the same file. This checks the efficiency of multi-process fsyncs
* against the same file. Possibly this should be done with writethrough
* on platforms which support it.
pg_test_fsync is intended to give you a reasonable idea of what the fastest wal_sync_method is on your specific system, as well as supplying diagnostic information in the event of an identified I/O problem. However, differences shown by pg_test_fsync might not make any difference in real database throughput, especially since many database servers are not speed-limited by their transaction logs.
源代码在$SRC/contrib/pg_test_fsync
例一
在一个sas 10k raid1的磁盘组 , ext3文件系统下的测试 .
postgres@digoal-> pg_test_fsync
2000 operations per test
O_DIRECT supported on this platform for open_datasync and open_sync.
Compare file sync methods using one 8kB write:
(in wal_sync_method preference order, except fdatasync
is Linux's default)
open_datasync n/a
fdatasync 159.689 ops/sec
fsync 161.249 ops/sec
fsync_writethrough n/a
open_sync 162.821 ops/sec
Compare file sync methods using two 8kB writes:
(in wal_sync_method preference order, except fdatasync
is Linux's default)
open_datasync n/a
fdatasync 160.920 ops/sec
fsync 161.076 ops/sec
fsync_writethrough n/a
open_sync 81.102 ops/sec
Compare open_sync with different write sizes:
(This is designed to compare the cost of writing 16kB
in different write open_sync sizes.)
16kB open_sync write 163.539 ops/sec
8kB open_sync writes 81.464 ops/sec
4kB open_sync writes 40.652 ops/sec
2kB open_sync writes 19.960 ops/sec
1kB open_sync writes 10.130 ops/sec
Test if fsync on non-write file descriptor is honored:
(If the times are similar, fsync() can sync data written
on a different descriptor.) 以下数值类似的话，则表示支持multi-processes fsync同一个文件
write, fsync, close 157.128 ops/sec
write, close, fsync 152.587 ops/sec
Non-Sync'ed 8kB writes:
write 170386.778 ops/sec
n/a表示此系统不支持此sync method, 如果设置为不支持的WAL_SYNC_METHOD，启动数据库时将报错，
如下 :
postgres@digoal-> FATAL: invalid value for parameter "wal_sync_method": "fsync_writethrough"
HINT: Available values: fsync, fdatasync, open_sync.
例二:
在RAMDISK上测试 fsync的性能 ,
postgres@digoal-> cd /dev/shm
postgres@digoal-> pg_test_fsync
2000 operations per test
O_DIRECT supported on this platform for open_datasync and open_sync.
Compare file sync methods using one 8kB write:
(in wal_sync_method preference order, except fdatasync
is Linux's default)
open_datasync n/a
fdatasync 412626.367 ops/sec
fsync 409081.612 ops/sec
fsync_writethrough n/a
open_sync n/a*
* This file system and its mount options do not support direct
I/O, e.g. ext4 in journaled mode.
Compare file sync methods using two 8kB writes:
(in wal_sync_method preference order, except fdatasync
is Linux's default)
open_datasync n/a
fdatasync 232369.002 ops/sec
fsync 234000.234 ops/sec
fsync_writethrough n/a
open_sync n/a*
* This file system and its mount options do not support direct
I/O, e.g. ext4 in journaled mode.
Compare open_sync with different write sizes:
(This is designed to compare the cost of writing 16kB
in different write open_sync sizes.)
16kB open_sync write n/a*
8kB open_sync writes n/a*
4kB open_sync writes n/a*
2kB open_sync writes n/a*
1kB open_sync writes n/a*
Test if fsync on non-write file descriptor is honored:
(If the times are similar, fsync() can sync data written
on a different descriptor.)
write, fsync, close 155557.284 ops/sec
write, close, fsync 161864.681 ops/sec
Non-Sync'ed 8kB writes:
write 282445.982 ops/sec