windows - Redis 集群中过于详细的日志记录

Question

我有一个具有以下 redis.conf 的 Redis 集群设置:

bind 127.0.0.1
bind 10.0.0.20

port 7000
cluster-enabled yes
cluster-config-file nodes.7000.conf
cluster-node-timeout 15000
appendonly yes
appendfilename "appendonly.7000.aof"

timeout 0
tcp-keepalive 60

databases 1
stop-writes-on-bgsave-error no

loglevel warning
logfile "serverlog.7000.txt"

集群中的每台机器都有2个Redis实例在运行；一个在 7000 端口，一个在 7001 端口。

即使我的日志级别是 warning，我仍然每秒收到几条日志语句。 serverlog.7000.txt 看起来像这样:

[1112] 09 Apr 06:44:28.623 # WSA_IO_PENDING writing to socket fd 14
[1112] 09 Apr 06:44:28.623 # clusterWriteDone written 2520 fd 14
[1112] 09 Apr 06:44:28.842 # WSA_IO_PENDING writing to socket fd 27
[1112] 09 Apr 06:44:28.842 # clusterWriteDone written 2520 fd 27
[1112] 09 Apr 06:44:28.936 # WSA_IO_PENDING writing to socket fd 23
[1112] 09 Apr 06:44:28.936 # clusterWriteDone written 2520 fd 23
[1112] 09 Apr 06:44:29.717 # WSA_IO_PENDING writing to socket fd 17
[1112] 09 Apr 06:44:29.717 # clusterWriteDone written 2520 fd 17
[1112] 09 Apr 06:44:29.748 # WSA_IO_PENDING writing to socket fd 21
[1112] 09 Apr 06:44:29.748 # clusterWriteDone written 2520 fd 21
[1112] 09 Apr 06:44:30.607 # WSA_IO_PENDING writing to socket fd 10
[1112] 09 Apr 06:44:30.607 # clusterWriteDone written 2520 fd 10
[1112] 09 Apr 06:44:30.811 # WSA_IO_PENDING writing to socket fd 11
[1112] 09 Apr 06:44:30.811 # clusterWriteDone written 2520 fd 11
[1112] 09 Apr 06:44:31.795 # WSA_IO_PENDING writing to socket fd 13
[1112] 09 Apr 06:44:31.795 # clusterWriteDone written 2520 fd 13
[1112] 09 Apr 06:44:31.904 # WSA_IO_PENDING writing to socket fd 12
[1112] 09 Apr 06:44:31.904 # clusterWriteDone written 2416 fd 12
[1112] 09 Apr 06:44:32.217 # WSA_IO_PENDING writing to socket fd 23
[1112] 09 Apr 06:44:32.217 # clusterWriteDone written 2520 fd 23
[1112] 09 Apr 06:44:32.779 # WSA_IO_PENDING writing to socket fd 16
[1112] 09 Apr 06:44:32.779 # clusterWriteDone written 2520 fd 16
[1112] 09 Apr 06:44:32.858 # WSA_IO_PENDING writing to socket fd 29
[1112] 09 Apr 06:44:32.858 # clusterWriteDone written 2520 fd 29
[1112] 09 Apr 06:44:32.998 # WSA_IO_PENDING writing to socket fd 11
[1112] 09 Apr 06:44:32.998 # clusterWriteDone written 2520 fd 11
[1112] 09 Apr 06:44:34.092 # WSA_IO_PENDING writing to socket fd 17
[1112] 09 Apr 06:44:34.092 # clusterWriteDone written 2520 fd 17
[1112] 09 Apr 06:44:35.186 # WSA_IO_PENDING writing to socket fd 14
[1112] 09 Apr 06:44:35.186 # clusterWriteDone written 2520 fd 14
[1112] 09 Apr 06:44:35.186 # WSA_IO_PENDING writing to socket fd 117
[1112] 09 Apr 06:44:35.186 # clusterWriteDone written 2416 fd 117
[1112] 09 Apr 06:44:35.404 # WSA_IO_PENDING writing to socket fd 27
[1112] 09 Apr 06:44:35.404 # clusterWriteDone written 2520 fd 27
[1112] 09 Apr 06:44:35.483 # WSA_IO_PENDING writing to socket fd 19
[1112] 09 Apr 06:44:35.483 # clusterWriteDone written 2520 fd 19
[1112] 09 Apr 06:44:36.076 # WSA_IO_PENDING writing to socket fd 10
[1112] 09 Apr 06:44:36.076 # clusterWriteDone written 2520 fd 10
[1112] 09 Apr 06:44:36.279 # WSA_IO_PENDING writing to socket fd 13
[1112] 09 Apr 06:44:36.279 # clusterWriteDone written 2520 fd 13
[1112] 09 Apr 06:44:37.295 # WSA_IO_PENDING writing to socket fd 21
[1112] 09 Apr 06:44:37.295 # clusterWriteDone written 2520 fd 21
[1112] 09 Apr 06:44:37.373 # WSA_IO_PENDING writing to socket fd 12
[1112] 09 Apr 06:44:37.373 # clusterWriteDone written 2520 fd 12
[1112] 09 Apr 06:44:37.686 # WSA_IO_PENDING writing to socket fd 23
[1112] 09 Apr 06:44:37.686 # clusterWriteDone written 2520 fd 23

这每天都会创建几百兆的日志数据，这有点难以完成，也需要我们更仔细地观察磁盘空间。对我来说，这个日志事件看起来不像警告。有什么方法可以关闭它们或解决问题吗？我已经尝试了所有其他记录的日志级别，但似乎没有什么能让它们消失。

我在 Windows 2012 上使用 Redis 3.2.100。

更新:我注意到这些也完全淹没了 Windows 事件日志。

Answer 1

在 3.2.100 中无法专门关闭这些“警告”，因为日志语句使用最高的 LL_WARNING 级别。这个问题已在我未维护的 MSOpenTech 存储库(我将其更新为 Redis 4.0.2)的分支中报告，并且已通过将该级别降低到 LL_DEBUG 来解决。更多详情:https://github.com/tporadowski/redis/issues/14

此更改将包含在下一个版本 (4.0.2.3) 中，或者您可以获取最新的源代码并自行构建。

可在此处找到当前版本:https://github.com/tporadowski/redis/releases