pg运维之REDUCE WAL BY INCREASING CHECKPOINT DISTANCE

如果你在生产中运行PostgreSQL，你可能已经注意到，调整检查点对调整服务器和提高数据库整体性能是一件非常有益的事情。然而，这里面还有更多的内容。增加检查点的距离也可以帮助实际减少首先创建的WAL的数量。这是对的。这不仅仅是一个性能问题，而且对创建的日志量也有影响。

Creating a test installation

设置参数

max_wal_size = 128MB
synchronous_commit = off

写入测试数据

% createdb test
% pgbench -s 100 -i test
dropping old tables...
NOTICE:  table "pgbench_accounts" does not exist, skipping
NOTICE:  table "pgbench_branches" does not exist, skipping
NOTICE:  table "pgbench_history" does not exist, skipping
NOTICE:  table "pgbench_tellers" does not exist, skipping
creating tables...
generating data...
100000 of 10000000 tuples (1%) done (elapsed 0.13 s, remaining 12.90 s)
200000 of 10000000 tuples (2%) done (elapsed 0.28 s, remaining 13.79 s)
…
9700000 of 10000000 tuples (97%) done (elapsed 12.32 s, remaining 0.38 s)
9800000 of 10000000 tuples (98%) done (elapsed 12.46 s, remaining 0.25 s)
9900000 of 10000000 tuples (99%) done (elapsed 12.59 s, remaining 0.13 s)
10000000 of 10000000 tuples (100%) done (elapsed 12.68 s, remaining 0.00 s)
vacuuming...
creating primary keys...
done.

running a PostgreSQL benchmark

现在让我们运行一个测试。我们想运行10×100万个交易。如果你想在你的数据库上重复测试，确保测试运行的时间足够长。让我们来看看这个测试!

% pgbench -c 10 -j 5 -t 1000000 test
starting vacuum...end.
transaction type: <builtin: TPC-B (sort of)>
scaling factor: 100
query mode: simple
number of clients: 10
number of threads: 5
number of transactions per client: 1000000
number of transactions actually processed: 10000000/10000000
latency average = 0.730 ms
tps = 13699.655677 (including connections establishing)
tps = 13699.737288 (excluding connections establishing)

正如你所看到的，我们成功地运行了大约每秒13.700个交易（在我的个人笔记本电脑上）。让我们来看看创建的WAL的数量。

test=# SELECT pg_current_wal_insert_lsn();
pg_current_wal_insert_lsn
---------------------------
13/522921E8
(1 row)

计算出当前WAL位置的方法是调用pg_current_insert_lsn()函数。对于大多数人来说，这种格式可能是无法理解的，因此将该值转化为可读的数字比较容易。

test=# SELECT pg_size_pretty(pg_current_wal_insert_lsn() - '0/00000000'::pg_lsn);
pg_size_pretty
----------------
77 GB
(1 row)

Increasing checkpoint distances

设置参数

max_wal_size = 150GB
checkpoint_timeout = 1d
synchronous_commit = off

test=# SELECT pg_current_wal_insert_lsn();
pg_current_wal_insert_lsn
---------------------------
1/4F210328
(1 row)test=# SELECT pg_size_pretty(pg_current_wal_insert_lsn() - '0/00000000'::pg_lsn);
pg_size_pretty
----------------
5362 MB
(1 row)

仅仅产生了5.3G的wal