标签：03 set 17 .. overcommit Apr 52 2023 2685
问题描述：启动redis时，告警日志中出现“The TCP backlog setting of 511……”以及“overcommit_memory is set to 0…..”警告，如下所示：
系统：rhel 7.9
数据库：redis 6.2.6
1、异常重现
[root@leo-redis626-a redis-6.2.6]# redis-server redis6379.conf
[root@leo-redis626-a redis-6.2.6]# tail -5000f redis.log
2685:C 03 Apr 2023 17:52:14.825 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
2685:C 03 Apr 2023 17:52:14.825 # Redis version=6.2.6, bits=64, commit=00000000, modified=0, pid=2685, just started
2685:C 03 Apr 2023 17:52:14.825 # Configuration loaded
2685:M 03 Apr 2023 17:52:14.829 * Increased maximum number of open files to 10032 (it was originally set to 1024).
2685:M 03 Apr 2023 17:52:14.829 * monotonic clock: POSIX clock_gettime
2685:M 03 Apr 2023 17:52:14.830 * Running mode=standalone, port=6379.
2685:M 03 Apr 2023 17:52:14.831 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.
2685:M 03 Apr 2023 17:52:14.831 # Server initialized
2685:M 03 Apr 2023 17:52:14.831 # WARNING overcommit_memory is set to 0! Background save may fail under low memory condition. To fix this issue add 'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command 'sysctl vm.overcommit_memory=1' for this to take effect.
2685:M 03 Apr 2023 17:52:14.831 * Loading RDB produced by version 6.2.6
2685:M 03 Apr 2023 17:52:14.831 * RDB age 962 seconds
2685:M 03 Apr 2023 17:52:14.832 * RDB memory usage when created 2.05 Mb
2685:M 03 Apr 2023 17:52:14.832 # Done loading RDB, keys loaded: 1, keys expired: 0.
2685:M 03 Apr 2023 17:52:14.832 * DB loaded from disk: 0.001 seconds
2685:M 03 Apr 2023 17:52:14.832 * Ready to accept connections
2、解决过程
2.1、警告1
警告原因：该警告表示无法强制设置TCP backlog为511，因为/proc/sys/net/core/somaxconn设置为较低的值128。
解决方法：
A、echo 511 > /proc/sys/net/core/somaxconn，注意此命令只是暂时生效，重启后会失效。
B、编辑/etc/sysctl.conf，添加net.core.somaxconn = 1024，然后执行sysctl -p使其永久生效。
实施过程：
[root@leo-redis626-a redis-6.2.6]# cat <<EOF>>/etc/sysctl.conf
net.core.somaxconn = 1024
EOF
[root@leo-redis626-a ~]# /sbin/sysctl -p
net.core.somaxconn = 1024
2.2、警告2
警告原因：
overcommit_memory设置为0，在内存不足的情况下，后台保存可能失败。
解决方法：
--执行命令进行配置
[root@leo-redis626-a redis-6.2.6]# echo 1 > /proc/sys/vm/overcommit_memory
说明：此后重启redis，告警日志不在报错.
相关值的介绍：
0：表示内核将检查是否有足够的可用内存供应用进程使用；如果有足够的可用内存，内存申请允许；否则内存申请失败，并把错误返回给应用进程。
1：表示内核允许分配所有的物理内存，而不管当前的内存状态如何。
2：表示内核允许分配超过所有物理内存和交换空间总和的内存。
参考网址：
https://blog.csdn.net/afreon/article/details/118727725
https://blog.csdn.net/u013131716/article/details/118938471
标签：03,
set,
17,
..,
overcommit,
Apr,
52,
2023,
2685
From： https://blog.51cto.com/u_12991611/6167501