After checking DHCR prod container I found that container had 39 restarts in last 4 days.
Restarts are caused with the error 137 which means that container was out of memory (OOM).
Since all three namespaces were jailed to 128Mb of RAM, I extended it now to 1Gb per namespace.
Please redeploy the prod container that it can use more RAM that it now available.
If this continues to happen the app probably has memory leak.
Relevant quotes (not written by me):