I run foundationdb in a pod on ocp. I notice that if I delete the pod and let kubernetes to recreate it, the old log files that belongs to the old pod (with a different ip) in fdb-trace-logs didn’t get removed and a new set of logs are created for the new pod with the new ip. Even rebooting the ocp cluster that would not help and that will eventually eat up my ephermal-storage overtime although each process has 100MiB limit, but if the cluster/pod get rebooted everyday, this will accumulate. Is there a setting that I can use to instruct fdbserver to not keep those old ip logs or only keep 1 set at most?
I believe that on the fdbserver side there is no workaround for this (at least I’m not aware of any) for the client you should be able to set the trace file identifier, e.g. foundationdb/fdb.options at main · apple/foundationdb · GitHub it might be that this. works for the fdbserver process but I haven’t tested this yet.