Constant Out Of Memory spikes since last v. 10.11.17

Happily ran a cluster on the same config for 3 years but since updating to 10.11.17 I experiance out of memory killer after sudden spikes. There seems to be no query, time or commonality to the issue. The container itself is 24Gb of RAM the process sits at 11Gb or RAM all day then whacks the stops at 24Gb for about a minute and then gets terminated by OOM killer.

Lots of the same.

I have attempted over the years to make best use of RAM but I have basically kept halving any settings as well as increasing RAM available but there is something definitely going wrong and breaking it. Winding back to 16 is so difficult that its more functionally impossible.

Jun 1 16:41:20 prox-webcdb-01 systemd[1]: mariadb.service: Failed with result 'oom-kill'.
Jun 1 16:43:51 prox-webcdb-01 kernel: [43981016.526271] oom-kill:constraint=CONSTRAINT_MEMCG,nodemask=(null),cpuset=ns,mems_allowed=0-1,oom_memcg=/lxc/135,task_memcg=/lxc/135/ns/system.slice/mariadb.service,task=mariadbd,pid=2403993,uid=112
Jun 1 18:12:51 prox-webcdb-01 kernel: [43986356.312600] mariadbd invoked oom-killer: gfp_mask=0xcc0(GFP_KERNEL), order=0, oom_score_adj=-500
Jun 1 18:12:52 prox-webcdb-01 systemd[1]: mariadb.service: Failed with result 'oom-kill'.
Jun 3 15:05:03 prox-webcdb-01 systemd[1]: mariadb.service: Failed with result 'oom-kill'.