PostgreSQL 源码性能诊断(perf profiling)指南 - 珍藏级
背景
数据库的性能优化是一个非常经典的话题，数据库的优化手段以及优化的角度也各不相同。
例如，可以从OS内核、网络、块设备、编译器、文件系统、SQL、数据库参数、业务逻辑、源码等各个方面去进行优化。
但是如果在优化前了解瓶颈在什么地方，可以向打鼹鼠一样，先打大的，起到事半功倍的效果。
本文将针对Linux平台下的PostgreSQL服务，讲解如何诊断PostgreSQL 数据库服务的瓶颈，了解瓶颈在哪个方面，甚至可以了解到瓶颈在哪段代码？
有了比较详细的性能诊断报告，可以给出比较好的指导优化建议。
我们要用到的工具是 《Linux 性能诊断(profiling) perf使用指南》
使用perf跟踪PostgreSQL的前提
要输出完备的跟踪信息，如符号表，call stack traces, 汇编指令。必须在编译PostgreSQL时设置相应的编译开关，例如GCC需要开启如下
CFLAGS="-g -ggdb -fno-omit-frame-pointer"
建议Linux内核编译时加上
CONFIG_KALLSYMS=y
CONFIG_FRAME_POINTER=y
编译perf时需要支持libunwind, 并加上
gcc -g dwarf
man gcc
-gdwarf-version
Produce debugging information in DWARF format (if that is supported). This is the format used by DBX on IRIX 6. The value of version may be either 2 or 3; the default version is 3.
Note that with DWARF version 2 some ports require, and will always use, some non-conflicting DWARF 3 extensions in the unwind tables.
如果是yum安装的软件，可以安装对应的debuginfo包。
例子
wget https://ftp.postgresql.org/pub/source/v9.6.1/postgresql-9.6.1.tar.bz2
tar -jxvf postgresql-9.6.1.tar.bz2
cd postgresql-9.6.1
export USE_NAMED_POSIX_SEMAPHORES=1
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto -g -ggdb -fno-omit-frame-pointer" ./configure --prefix=/home/digoal/pgsql9.6
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto -g -ggdb -fno-omit-frame-pointer" make world -j 64
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto -g -ggdb -fno-omit-frame-pointer" make install-world
常用跟踪手段
1. benchmark
首先，需要定一个目标，比如insert only的场景，如何找到其瓶颈。
创建测试表
postgres=# create table perf_test_ins(id serial8 primary key, info text, crt_time timestamp);
alCREATE TABLE
postgres=# alter sequence perf_test_ins_id_seq cache 10000;
ALTER SEQUENCE
创建测试脚本，不停插入
vi test.sql
insert into perf_test_ins(info,crt_time) values ('test',now());
疲劳压测开始
pgbench -M prepared -n -r -P 1 -f ./test.sql -h $PGDATA -c 32 -j 32 -T 600
2. 实时跟踪
pgbench压测的同时，实时观测。
su - root
perf top -agv -F 1000
解释
perf top --help
NAME
perf-top - System profiling tool.
SYNOPSIS
perf top [-e <EVENT> | --event=EVENT] [<options>]
DESCRIPTION
This command generates and displays a performance counter profile in real time.
-F <freq>, --freq=<freq>
Profile at this frequency.
-a, --all-cpus
System-wide collection. (default)
-g, --group
Put the counters into a counter group.
-K, --hide_kernel_symbols
Hide kernel symbols.
-U, --hide_user_symbols
Hide user symbols.
-v, --verbose
Be more verbose (show counter open errors, etc).
输出
PerfTop: 23287 irqs/sec kernel:31.2% exact: 0.0% [1000Hz cycles], (all, 24 CPUs)
--------------------------------------------------------------------------------------------------------------
samples pcnt RIP function DSO
_______ _____ ________________ _________________________________________ _____________________________________
4179.00 1.6% 0000000000524e10 AllocSetAlloc.lto_priv.1186 /home/digoal/pgsql9.6/bin/postgres
3966.00 1.5% 0000000000043d10 __GI_vfprintf /lib64/libc-2.12.so
3822.00 1.5% ffffffff8150bb50 schedule [kernel.kallsyms]
3659.00 1.4% 0000000000082e10 __memset_sse2 /lib64/libc-2.12.so
3523.00 1.4% 00000000005782d0 LWLockAcquire.constprop.931 /home/digoal/pgsql9.6/bin/postgres
3219.00 1.2% 00000000002c9390 LWLockRelease /home/digoal/pgsql9.6/bin/postgres
3114.00 1.2% 0000000000332e10 PostgresMain /home/digoal/pgsql9.6/bin/postgres
3042.00 1.2% ffffffff8150e8b0 _spin_lock [kernel.kallsyms]
2948.00 1.1% 0000000000573fe0 hash_search_with_hash_value.constprop.930 /home/digoal/pgsql9.6/bin/postgres
2901.00 1.1% 00000000005746d0 hash_search_with_hash_value.constprop.928 /home/digoal/pgsql9.6/bin/postgres
2848.00 1.1% 00000000002db5f0 GetSnapshotData /home/digoal/pgsql9.6/bin/postgres
2471.00 1.0% 00000000005338b0 hash_seq_search /home/digoal/pgsql9.6/bin/postgres
2275.00 0.9% 0000000000579ad0 LWLockAcquire.constprop.932 /home/digoal/pgsql9.6/bin/postgres
2244.00 0.9% 0000000000078460 _int_malloc /lib64/libc-2.12.so
2215.00 0.9% 0000000000088aa0 memcpy /lib64/libc-2.12.so
2163.00 0.8% 00000000002df050 LockAcquireExtended /home/digoal/pgsql9.6/bin/postgres
2057.00 0.8% 0000000000131a70 __strlen_sse42 /lib64/libc-2.12.so
2017.00 0.8% 00000000002ba2d0 _bt_compare /home/digoal/pgsql9.6/bin/postgres
1977.00 0.8% 0000000000006e40 doCustom.lto_priv.13 /home/digoal/pgsql9.6/bin/pgbench
1951.00 0.8% ffffffff8100a630 __switch_to [kernel.kallsyms]
1855.00 0.7% 000000000053d5e0 hash_search_with_hash_value /home/digoal/pgsql9.6/bin/postgres
1767.00 0.7% 00000000001c6670 hash_any /home/digoal/pgsql9.6/bin/postgres
1726.00 0.7% 00000000002c8310 PinBuffer.isra.3.lto_priv.2340 /home/digoal/pgsql9.6/bin/postgres
1668.00 0.6% 00000000000ee570 XLogInsertRecord /home/digoal/pgsql9.6/bin/postgres
1604.00 0.6% 0000000000250500 pgstat_report_activity /home/digoal/pgsql9.6/bin/postgres
1572.00 0.6% ffffffff81189c00 fget_light [kernel.kallsyms]
1565.00 0.6% 0000000000426970 SearchCatCache /home/digoal/pgsql9.6/bin/postgres
1460.00 0.6% ffffffff810e03a0 __audit_syscall_exit [kernel.kallsyms]
1415.00 0.5% ffffffff81057e80 update_curr [kernel.kallsyms]
1409.00 0.5% 00000000000777f0 _int_free /lib64/libc-2.12.so
1386.00 0.5% 00000000000fbc10 CommitTransaction.lto_priv.2946 /home/digoal/pgsql9.6/bin/postgres
1296.00 0.5% ffffffff8119f6d0 do_select [kernel.kallsyms]
1285.00 0.5% 00000000002c8690 LWLockWaitListLock.lto_priv.1265 /home/digoal/pgsql9.6/bin/postgres
1268.00 0.5% ffffffff8150e790 _spin_lock_irqsave [kernel.kallsyms]
1247.00 0.5% 0000000000577650 hash_search_with_hash_value.constprop.929 /home/digoal/pgsql9.6/bin/postgres
1220.00 0.5% 0000000000538a10 AllocSetFree.lto_priv.1187 /home/digoal/pgsql9.6/bin/postgres
1201.00 0.5% 0000000000389660 ExecModifyTable /home/digoal/pgsql9.6/bin/postgres
1185.00 0.5% 000000000009b6d0 gettimeofday /lib64/libc-2.12.so
1171.00 0.5% ffffffff81278a60 copy_user_generic_string [kernel.kallsyms]
1142.00 0.4% 000000000013cb10 __memcmp_sse4_1 /lib64/libc-2.12.so
1124.00 0.4% 0000000000032620 __GI___sigsetjmp /lib64/libc-2.12.so
1123.00 0.4% 00000000000ef2b0 XLogInsert /home/digoal/pgsql9.6/bin/postgres
1112.00 0.4% 00000000002d29e0 LockReleaseAll /home/digoal/pgsql9.6/bin/postgres
1108.00 0.4% 00000000003c7620 ExecInitExpr /home/digoal/pgsql9.6/bin/postgres
1056.00 0.4% 000000000000ac80 threadRun.lto_priv.20 /home/digoal/pgsql9.6/bin/pgbench
1052.00 0.4% ffffffff81062a40 select_task_rq_fair [kernel.kallsyms]
1033.00 0.4% 0000000000017710 pqParseInput3 /home/digoal/pgsql9.6/lib/libpq.so.5.9
按Enter键弹出可以配置的选项以及当前状态。
Mapped keys:
[d] display refresh delay. (2)
[e] display entries (lines). (47)
[f] profile display filter (count). (5)
[F] annotate display filter (percent). (5%)
[s] annotate symbol. (NULL)
[S] stop annotation.
[K] hide kernel_symbols symbols. (no) 是否显示kernel symbols
[U] hide user symbols. (no) 是否显示user symbols
[z] toggle sample zeroing. (0)
[qQ] quit.
调整压测命令，连接回环地址，(前面连接的是unix socket)
pgbench -M prepared -n -r -P 1 -f ./test.sql -h 127.0.0.1 -c 32 -j 32 -T 600
再次观测perf top结果, 采样消耗排名第一的变成了这条，找到了一个瓶颈，关闭iptables，这条就消失了，TPS也提升了。
5871.00 1.7% 0000000000002e40 ipt_do_table /lib/modules/2.6.32-358.23.2.ali1195.el6.x86_64/kernel/net/ipv4/netfilter/ip_tables.ko
3. 跟踪事件计数（指定事件）
指定命令进行跟踪，或者指定PID进行跟踪。
NAME
perf-stat - Run a command and gather performance counter statistics
SYNOPSIS
perf stat [-e <EVENT> | --event=EVENT] [-a] <command>
perf stat [-e <EVENT> | --event=EVENT] [-a] — <command> [<options>]
DESCRIPTION
This command runs a command and gathers performance counter statistics from it.
-a, --all-cpus
system-wide collection from all CPUs
-c, --scale
scale/normalize counter values
-B, --big-num
print large numbers with thousands′ separators according to locale
-v, --verbose
be more verbose (show counter open errors, etc)
-p, --pid=<pid>
stat events on existing process id
-t, --tid=<tid>
stat events on existing thread id
-r, --repeat=<n>
repeat command and print average + stddev (max: 100)
跟踪批量插入一批记录的事件调用统计
event来自perf list的输出
perf stat -e block:*,syscalls:*,xfs:* -a -B -v -c -- /home/digoal/pgsql9.6/bin/psql -h 127.0.0.1 -p 5288 -q -U postgres postgres -c "drop table if exists x; create table x as select a FROM generate_series(1,1000000) a;";
输出
0 block:block_rq_abort [98.79%]
0 block:block_rq_requeue [98.79%]
995 block:block_rq_complete [98.79%]
993 block:block_rq_insert [98.79%]
997 block:block_rq_issue [98.79%]
0 block:block_bio_bounce [98.79%]
0 block:block_bio_complete [98.79%]
18 block:block_bio_backmerge [98.79%]
1 block:block_bio_frontmerge [98.79%]
1,011 block:block_bio_queue [98.79%]
992 block:block_getrq [98.79%]
2 block:block_sleeprq [98.79%]
250 block:block_plug [98.79%]
2 block:block_unplug_timer [98.79%]
361 block:block_unplug_io [98.79%]
0 block:block_split [98.79%]
1,011 block:block_remap [98.79%]
0 block:block_rq_remap [98.79%]
5 syscalls:sys_enter_socket [98.79%]
5 syscalls:sys_exit_socket [98.79%]
0 syscalls:sys_enter_socketpair [98.79%]
0 syscalls:sys_exit_socketpair [98.79%]
... 略 ...
计数例子
perf stat -a -B -v -c -- /home/digoal/pgsql9.6/bin/psql -h 127.0.0.1 -p 5288 -q -U postgres postgres -c "drop table if exists x; create table x as select a FROM generate_series(1,1000000) a;";
22433.920027 task-clock # 24.042 CPUs utilized [100.00%]
4,275 context-switches # 0.000 M/sec [100.00%]
98 CPU-migrations # 0.000 M/sec [100.00%]
56,484 page-faults # 0.003 M/sec
2,687,623,473 cycles # 0.120 GHz [80.73%]
2,747,044,652 stalled-cycles-frontend # 102.21% frontend cycles idle [92.42%]
2,358,755,489 stalled-cycles-backend # 87.76% backend cycles idle [81.89%]
3,373,693,967 instructions # 1.26 insns per cycle 越高越好，说明CPU没闲着。至少要大于1，但是注意同步锁等待这个值也可能很高，CPU空转，这就不好了。
# 0.81 stalled cycles per insn [91.50%] 越低越好，说明CPU没有闲着。
632,351,270 branches # 28.187 M/sec [91.82%]
1,236,798 branch-misses # 0.20% of all branches [53.19%]
0.933128299 seconds time elapsed
perf stat重点看的是instructions，判断CPU是否充分被利用，如果没有，说明系统有其他瓶颈，例如IO等待，或者其他等待。
4. 采样跟踪
这个是用得最多的，先采样，后分析报告。
NAME
perf-record - Run a command and record its profile into perf.data
SYNOPSIS
perf record [-e <EVENT> | --event=EVENT] [-l] [-a] <command>
perf record [-e <EVENT> | --event=EVENT] [-l] [-a] — <command> [<options>]
DESCRIPTION
This command runs a command and gathers a performance counter profile from it, into perf.data - without displaying anything.
This file can then be inspected later on, using perf report.
-A, --append
Append to the output file to do incremental profiling.
-o, --output=
Output file name.
-e, --event= event来自perf list的输出
Select the PMU event.
Selection can be a symbolic event name (use perf list to list all events) or a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a hexadecimal event descriptor.
-a, --all-cpus
System-wide collection from all CPUs.
-g, --call-graph
Do call-graph (stack chain/backtrace) recording.
-v, --verbose
Be more verbose (show counter open errors, etc).
-s, --stat 相对的是默认的per cpu mode
Per thread counts.
-G name,..., --cgroup name,... 如果你只想跟踪某一些进程的统计信息，可以将这些进程放到一个cgroup中，跟踪这个cgroup的。
monitor only in the container (cgroup) called "name".
This option is available only in per-cpu mode.
The cgroup filesystem must be mounted.
All threads belonging to container "name" are monitored when they run on the monitored CPUs.
Multiple cgroups can be provided.
Each cgroup is applied to the corresponding event, i.e., first cgroup to first event, second cgroup to second event and so on. It is
possible to provide an empty cgroup (monitor all the time) using, e.g., -G foo,,bar.
Cgroups must have corresponding events, i.e., they always refer to events defined earlier on the command line.
跑pgbench压测，和前面的pgbench一样，不断并发插入。
然后开启统计收集，收集10秒的统计信息，默认输出到perf.data文件。
perf record -avg -- sleep 10
[ perf record: Woken up 142 times to write data ]
[ perf record: Captured and wrote 37.111 MB perf.data (~1621415 samples) ]
5. 生成报告
前面使用perf record收集了统计信息到perf.data，接下来就对perf.data进行分析，输出报告。
NAME
perf-report - Read perf.data (created by perf record) and display the profile
SYNOPSIS
perf report [-i <file> | --input=file]
DESCRIPTION
This command displays the performance counter profile information recorded via perf record.
-i, --input=
Input file name. (default: perf.data)
-v, --verbose
Be more verbose. (show symbol address, etc)
-n, --show-nr-samples
Show the number of samples for each symbol
--showcpuutilization
Show sample percentage for different cpu modes.
-T, --threads
Show per-thread event counters
-U, --hide-unresolved
Only display entries resolved to a symbol.
-s, --sort=
Sort by key(s): pid, comm, dso, symbol, parent.
-g [type,min], --call-graph
Display call chains using type and min percent threshold. type can be either:
· flat: single column, linear exposure of call chains.
· graph: use a graph tree, displaying absolute overhead rates.
· fractal: like graph, but displays relative rates. Each branch of the tree is considered as a new profiled object.
Default: fractal,0.5.
--stdio
Use the stdio interface.
--tui
Use the TUI interface, that is integrated with annotate and allows zooming into DSOs or threads, among other features. Use of --tui requires a tty, if one is not present, as when piping to other
commands, the stdio interface is used.
–tui模式可以查看概览，支持交互式指令。
输入E全展开，输入C全收敛。 展开后可以使用键盘的上下键进行浏览。
[.] 表示userspace
[k] 表示kernel
1.39% 0.00% 0.01% 15 postgres /home/digoal/pgsql9.6/bin/postgres 0x524e3a d [.] AllocSetAlloc.lto_priv.1186
1.29% 0.00% 0.01% 22 postgres /lib64/libc-2.12.so 0x83127 B [.] __memset_sse2
1.25% 0.00% 0.00% 1 postgres /home/digoal/pgsql9.6/bin/postgres 0x57831d d [.] LWLockAcquire.constprop.931
1.24% 0.00% 0.05% 107 postgres /home/digoal/pgsql9.6/bin/postgres 0x574751 d [.] hash_search_with_hash_value.constprop.928
1.07% 0.00% 0.00% 2 postgres /home/digoal/pgsql9.6/bin/postgres 0x2c946f d [.] LWLockRelease
1.01% 0.00% 0.00% 1 postgres /home/digoal/pgsql9.6/bin/postgres 0x574035 d [.] hash_search_with_hash_value.constprop.930
0.98% 0.00% 0.03% 74 postgres /home/digoal/pgsql9.6/bin/postgres 0x334832 d [.] PostgresMain
0.94% 0.00% 0.02% 49 postgres /home/digoal/pgsql9.6/bin/postgres 0x2db87b d [.] GetSnapshotData
0.94% 0.00% 0.00% 1 postgres [kernel.kallsyms] 0xffffffff8150e8b1 k [k] _spin_lock
0.83% 0.00% 0.00% 1 postgres [kernel.kallsyms] 0xffffffff8150edb0 k [k] page_fault
0.81% 0.00% 0.00% 1 postgres /lib64/libc-2.12.so 0x88aa0 B [.] memcpy
0.77% 0.00% 0.22% 465 postgres /home/digoal/pgsql9.6/bin/postgres 0x533939 d [.] hash_seq_search
0.75% 0.00% 0.00% 1 postgres /lib64/libc-2.12.so 0x454ac B [.] __GI_vfprintf
0.75% 0.00% 0.03% 59 postgres /home/digoal/pgsql9.6/bin/postgres 0x579b1d d [.] LWLockAcquire.constprop.932
0.73% 0.00% 0.02% 42 postgres /home/digoal/pgsql9.6/bin/postgres 0x2df3ab d [.] LockAcquireExtended
0.67% 0.00% 0.00% 2 postgres /home/digoal/pgsql9.6/bin/postgres 0x1c66f9 d [.] hash_any
0.65% 0.00% 0.02% 49 postgres /home/digoal/pgsql9.6/bin/postgres 0x2ba33a d [.] _bt_compare
0.63% 0.00% 0.00% 3 postgres /home/digoal/pgsql9.6/bin/postgres 0xee632 d [.] XLogInsertRecord
0.62% 0.00% 0.00% 1 postgres /home/digoal/pgsql9.6/bin/postgres 0x53d5e1 d [.] hash_search_with_hash_value
0.59% 0.00% 0.59% 1318 pgbench /lib64/libc-2.12.so 0x47956 B [.] __GI_vfprintf
0.57% 0.00% 0.00% 1 postgres /lib64/libc-2.12.so 0x131aa7 B [.] __strlen_sse42
0.55% 0.00% 0.55% 1241 pgbench /home/digoal/pgsql9.6/bin/pgbench 0x85b3 d [.] doCustom.lto_priv.13
0.55% 0.00% 0.03% 55 postgres /home/digoal/pgsql9.6/bin/postgres 0x2c8327 d [.] PinBuffer.isra.3.lto_priv.2340
0.54% 0.00% 0.02% 35 postgres /home/digoal/pgsql9.6/bin/postgres 0x25059f d [.] pgstat_report_activity
0.53% 0.00% 0.02% 43 postgres /lib64/libc-2.12.so 0x78462 B [.] _int_malloc
0.52% 0.00% 0.01% 33 postgres /home/digoal/pgsql9.6/bin/postgres 0x426a36 d [.] SearchCatCache
0.47% 0.00% 0.00% 3 postgres [kernel.kallsyms] 0xffffffff81272a1c k [k] radix_tree_lookup_slot
0.45% 0.00% 0.01% 27 postgres /home/digoal/pgsql9.6/bin/postgres 0xfbe51 d [.] CommitTransaction.lto_priv.2946
0.44% 0.00% 0.00% 1 postgres /home/digoal/pgsql9.6/bin/postgres 0x5776f5 d [.] hash_search_with_hash_value.constprop.929
0.43% 0.00% 0.01% 30 postgres /home/digoal/pgsql9.6/bin/postgres 0x38a044 d [.] ExecModifyTable
0.41% 0.00% 0.00% 1 postgres [kernel.kallsyms] 0xffffffff8150bef4 k [k] schedule
0.40% 0.00% 0.01% 22 postgres /home/digoal/pgsql9.6/bin/postgres 0x2c86c2 d [.] LWLockWaitListLock.lto_priv.1265
0.40% 0.00% 0.01% 29 postgres /home/digoal/pgsql9.6/bin/postgres 0x2d2b2d d [.] LockReleaseAll
0.40% 0.00% 0.01% 12 postgres /home/digoal/pgsql9.6/bin/postgres 0x538a11 d [.] AllocSetFree.lto_priv.1187
0.39% 0.00% 0.00% 1 postgres [kernel.kallsyms] 0xffffffff8111e567 k [k] find_get_page
0.39% 0.01% 0.00% 24 postgres [kernel.kallsyms] 0xffffffff8148d1f8 k [k] tcp_sendmsg
0.39% 0.00% 0.01% 34 postgres /home/digoal/pgsql9.6/bin/postgres 0xef545 d [.] XLogInsert
0.39% 0.00% 0.01% 25 postgres /lib64/libc-2.12.so 0x32623 B [.] __GI___sigsetjmp
0.38% 0.00% 0.38% 860 pgbench /home/digoal/pgsql9.6/bin/pgbench 0xb3f3 d [.] threadRun.lto_priv.20
0.38% 0.00% 0.01% 33 postgres /lib64/libc-2.12.so 0x77cf2 B [.] _int_free
0.38% 0.00% 0.00% 1 postgres /lib64/libc-2.12.so 0x13cb10 B [.] __memcmp_sse4_1
0.38% 0.00% 0.02% 36 postgres /home/digoal/pgsql9.6/bin/postgres 0x3c762f d [.] ExecInitExpr
0.36% 0.36% 0.00% 816 pgbench [kernel.kallsyms] 0xffffffff8119f950 k [k] do_select
0.36% 0.00% 0.00% 4 postgres /home/digoal/pgsql9.6/bin/postgres 0x21e005 d [.] ReadBuffer_common.lto_priv.1215
0.36% 0.00% 0.01% 28 postgres /home/digoal/pgsql9.6/bin/postgres 0x1e5e50 d [.] _bt_doinsert
0.35% 0.00% 0.01% 16 postgres [vdso] 0x7fffb61ff8b0 ! [.] 0x7fffb61ff8b0
0.35% 0.00% 0.00% 1 postgres /lib64/libc-2.12.so 0x9b6db B [.] gettimeofday
0.35% 0.00% 0.35% 775 pgbench /home/digoal/pgsql9.6/lib/libpq.so.5.9 0x18057 d [.] pqParseInput3
0.32% 0.00% 0.01% 18 postgres /home/digoal/pgsql9.6/bin/postgres 0x2f9979 d [.] expression_tree_walker
0.31% 0.00% 0.01% 18 postgres /home/digoal/pgsql9.6/bin/postgres 0x306a5c d [.] socket_putmessage.lto_priv.1817
0.31% 0.31% 0.00% 702 pgbench [kernel.kallsyms] 0xffffffff8148c9be k [k] tcp_sendmsg
0.31% 0.00% 0.01% 17 postgres /home/digoal/pgsql9.6/bin/postgres 0x38891e d [.] ExecInitModifyTable
0.31% 0.00% 0.00% 1 postgres [kernel.kallsyms] 0xffffffff81057ff7 k [k] update_curr
0.31% 0.01% 0.00% 27 postgres [kernel.kallsyms] 0xffffffff81494ea6 k [k] tcp_ack
展开后的例子
- 1.39% 0.00% 0.01% 15 postgres /home/digoal/pgsql9.6/bin/postgres 0x524e3a d [.] AllocSetAlloc.lto_priv.1186
- AllocSetAlloc.lto_priv.1186
- 33.72% MemoryContextAlloc
- 27.57% MemoryContextCreate
- AllocSetContextCreate.constprop.955
- 43.77% StartTransaction.lto_priv.2177
start_xact_command.part.3
PostgresMain
ServerLoop
PostmasterMain
main
__libc_start_main
- 16.08% CreatePortal.constprop.203
PostgresMain
ServerLoop
PostmasterMain
main
__libc_start_main
- 14.06% MakePerTupleExprContext
ExecInsertIndexTuples.constprop.430
ExecModifyTable
ExecProcNode
standard_ExecutorRun
pgss_ExecutorRun
ProcessQuery
PortalRunMulti
PortalRun
PostgresMain
ServerLoop
PostmasterMain
main
__libc_start_main
- 13.35% CreateExecutorState
standard_ExecutorStart
pgss_ExecutorStart
ProcessQuery
PortalRunMulti
PortalRun
PostgresMain
ServerLoop
PostmasterMain
main
__libc_start_main
- 12.74% ExecInitResult
ExecInitNode
ExecInitModifyTable
ExecInitNode
standard_ExecutorStart
pgss_ExecutorStart
ProcessQuery
PortalRunMulti
PortalRun
PostgresMain
ServerLoop
...................略................
如果要查看详情，建议使用–stdio模式，看得非常清晰
#perf report -v -n --showcpuutilization -g --stdio
# Events: 224K cycles
#
# Overhead Samples sys us Command Shared Object
(全局开销占比) (采样) (系统开销占比, 系统开销耗时us) (命令) (共享对象)
# ........ .......... .............. ............................................... ........................................................
#
1.39% 0.00% 0.01% 15 postgres /home/digoal/pgsql9.6/bin/postgres 0x524e3a d [.] AllocSetAlloc.lto_priv.1186
|
--- AllocSetAlloc.lto_priv.1186
|
|--33.72%-- MemoryContextAlloc
| |
| |--27.57%-- MemoryContextCreate
| | AllocSetContextCreate.constprop.955
| | |
| | |--43.77%-- StartTransaction.lto_priv.2177
| | | start_xact_command.part.3
| | | PostgresMain
| | | ServerLoop
| | | PostmasterMain
| | | main
| | | __libc_start_main
| | |
| | |--16.08%-- CreatePortal.constprop.203
| | | PostgresMain
| | | ServerLoop
| | | PostmasterMain
| | | main
| | | __libc_start_main
| | |
| | |--14.06%-- MakePerTupleExprContext
| | | ExecInsertIndexTuples.constprop.430
| | | ExecModifyTable
| | | ExecProcNode
| | | standard_ExecutorRun
| | | pgss_ExecutorRun
| | | ProcessQuery
| | | PortalRunMulti
| | | PortalRun
| | | PostgresMain
| | | ServerLoop
| | | PostmasterMain
| | | main
| | | __libc_start_main
| | |
| | |--13.35%-- CreateExecutorState
| | | standard_ExecutorStart
| | | pgss_ExecutorStart
| | | ProcessQuery
| | | PortalRunMulti
| | | PortalRun
| | | PostgresMain
| | | ServerLoop
| | | PostmasterMain
| | | main
| | | __libc_start_main
...............略.............................
perf report输出为本文，可以较为详细的进行代码瓶颈的分析。
6. 生成火焰图
如果你觉得perf report的文本输出不够直观，可以生成火焰图，就更加直观了。
生成火焰图需要将perf record的原始数据弄出来，然后绘图。
perf.data的原始数据perf script即可输出。
生成火焰图
# git clone https://github.com/brendangregg/FlameGraph # or download it from github
# mv perf.data FlameGraph/
# cd FlameGraph
# perf script | ./stackcollapse-perf.pl > out.perf-folded
# cat out.perf-folded | ./flamegraph.pl > perf-kernel.svg
svg可以使用浏览器打开，转换的图片如下
指定内核内置事件跟踪
perf top|stat|record 都支持通过-e指定事件进行跟踪。
支持的事件如下
perf list
cpu-cycles OR cycles [Hardware event]
stalled-cycles-frontend OR idle-cycles-frontend [Hardware event]
stalled-cycles-backend OR idle-cycles-backend [Hardware event]
instructions [Hardware event]
cache-references [Hardware event]
cache-misses [Hardware event]
branch-instructions OR branches [Hardware event]
branch-misses [Hardware event]
bus-cycles [Hardware event]
......................略........................
事件支持写通配符，多个事件使用逗号隔开。
perf top -ag -e xfs:xfs_attr_list_sf,skb:kfree_skb,block:*,irq:softirq*
No CONFIG_PERF_EVENTS=y kernel support configured? 编译内核是需要配置CONFIG_PERF_EVENTS=y
perf stat -e xfs:xfs_attr_list_sf,skb:kfree_skb,block:*,irq:softirq* -- sleep 10
Performance counter stats for 'sleep 10':
0 xfs:xfs_attr_list_sf
0 skb:kfree_skb
0 block:block_rq_abort
0 block:block_rq_requeue
0 block:block_rq_complete
0 block:block_rq_insert
0 block:block_rq_issue
0 block:block_bio_bounce
0 block:block_bio_complete
0 block:block_bio_backmerge
0 block:block_bio_frontmerge
0 block:block_bio_queue
0 block:block_getrq
0 block:block_sleeprq
0 block:block_plug
0 block:block_unplug_timer
0 block:block_unplug_io
0 block:block_split
0 block:block_remap
0 block:block_rq_remap
0 irq:softirq_entry
0 irq:softirq_exit
0 irq:softirq_raise
10.001021697 seconds time elapsed
perf record -avg -e xfs:xfs_attr_list_sf,skb:kfree_skb,block:*,irq:softirq* -- sleep 10
perf report -vg --stdio
# Events: 17 skb:kfree_skb
#
# Overhead Command Shared Object Symbol
# ........ .............. ........................................... ................
#
35.29% init [kernel.kallsyms] 0xffffffff8103c94b k [k] native_safe_halt
|
--- native_safe_halt
11.76% syslog-ng /lib64/libpthread-2.12.so 0xea10 B [.] __recvmsg
|
--- __recvmsg
...........略.......................
kernel source 动态跟踪
如果perf list列出的事件不能满足跟踪需求，你可以自行添加任意代码进行跟踪。
这个功能需要perf和内核的支持，详见《Linux 性能诊断(profiling) perf使用指南》
1. Listing variables available for tcp_sendmsg():
# perf probe -V tcp_sendmsg
Available variables at tcp_sendmsg
@<tcp_sendmsg+0>
size_t size
struct kiocb* iocb
struct msghdr* msg
struct sock* sk
2. Creating a probe for tcp_sendmsg() with the "size" variable:
# perf probe --add 'tcp_sendmsg size'
Added new event:
probe:tcp_sendmsg (on tcp_sendmsg with size)
3. You can now use it in all perf tools, such as:
perf record -e probe:tcp_sendmsg -aR sleep 1
PostgreSQL source 动态跟踪
这个功能需要perf和软件编译时的支持，详见《Linux 性能诊断(profiling) perf使用指南》
sudo perf record -g dwarf -e block:block_rq_*,syscalls:sys_enter_write,syscalls:sys_enter_fsync -p 4495
sudo perf probe -x `which postgres` XLogFileInit
sudo perf probe -x `which postgres` XLogFileOpen
sudo perf probe -x `which postgres` XLogFlush
sudo perf probe -l
probe_postgres:XLogFileInit (on 0x000000000009a360)
probe_postgres:XLogFileOpen (on 0x000000000009a860)
probe_postgres:XLogFlush (on 0x00000000000a0670)
sudo perf record -g dwarf -u postgres -e probe_postgres:XLogFileInit,probe_postgres:XLogFileOpen,probe_postgres:XLogFlush
sudo perf stat -e probe_postgres:XLogFileInit,probe_postgres:XLogFileOpen,probe_postgres:XLogFlush -a -- /usr/pgsql-9.2/bin/pgbench -U postgres craig -c 2 -t 10000
starting vacuum...end.
transaction type: TPC-B (sort of)
scaling factor: 100
query mode: simple
number of clients: 2
number of threads: 1
number of transactions per client: 10000
number of transactions actually processed: 20000/20000
tps = 715.854663 (including connections establishing)
tps = 716.092133 (excluding connections establishing)
Performance counter stats for '/usr/pgsql-9.2/bin/pgbench -U postgres craig -c 2 -t 10000':
64 probe_postgres:XLogFileInit [100.00%]
0 probe_postgres:XLogFileOpen [100.00%]
55,440 probe_postgres:XLogFlush
27.987364469 seconds time elapsed
参考
1. https://wiki.postgresql.org/wiki/Profiling_with_perf
2. http://blog.2ndquadrant.com/tracing-postgresql-perf/
3. 《Linux 性能诊断(profiling) perf使用指南》
4. http://blog.2ndquadrant.com/testing-postgresql-vagrant-computer/
5. http://blog.2ndquadrant.com/devops-testing-postgresql-with-vagrant-on-your-computer-part-two/
6. https://www.quora.com/How-do-I-compile-a-Linux-perf-tool-with-all-features-For-Linux-4-0-on-Ubuntu