Service logs filled with 'ps' related error

[PaulVipond]

I'm getting lots of instances of this error in the logs for my postgresql service based on postgres:9.6.9-alpine:

data_postgresql.1.ft2gof5jci25@IS-57436 | ps: bad -o argument 'command', supported arguments: user,group,comm,args,pid,ppid,pgid,etime,nice,rgroup,ruser,time,tty,vsz,stat,rss

The errors aren't generated by any other containers using Postgres services, as they occur when the other containers aren't running. Googling around, I believe this may be fixed by using an enhanced version of 'ps' as follows because Alpine has a more limited version:

apk --no-cache add procps

Could this be included in future builds?

This might be related to an excess CPU consumption issue I've been having as the CPU is thrashed as it continually tries to execute the 'ps' command.

https://dba.stackexchange.com/questions/207589/excessive-postgres-docker-cpu-consumption

I haven't finished fully investigating the issue yet, so will update with my conclusions when done.

[tianon]

As noted on stack overflow, you've been hacked: https://dba.stackexchange.com/a/207636