We are using a 3 node galera cluster with Mysql 5.7 5.7.23-23-57-log on ubuntu 16.04 x64, 64G ram + 64g swap
my problem is that mysql memory usage is constantly growing and after a few days it runs to an OOM and got killed.
I’ve collected the following graphs from munin.
in the configuration I do not see any extrem value, we have already lowered them, no success.
Tried: update mysql, update OS, update kernel, reboot, half all the parameters, no luck
concurrent running user number is about 6, sleep is about 130 (connection pooling is in effect on the application)
I ran out of ideas, dunno how to troubleshoot this issue. thanks if you can point me to the right direction.
details from config:
key-buffer-size = 4M # 32M
max-allowed-packet = 12M
max-connect-errors = 1000000
tmp-table-size = 16M
max-heap-table-size = 16M
query-cache-type = 0
query-cache-size = 0
max-connections = 1000
thread-cache-size = 100
open-files-limit = 65535
table-definition-cache = 2048
table-open-cache = 10120
innodb-flush-method = O_DIRECT
innodb-log-files-in-group = 2
#zcs - half the orignal
innodb-log-file-size = 24M
innodb-flush-log-at-trx-commit = 1
innodb-file-per-table = 1
innodb-buffer-pool-size = 8G #16G