On Sat, Oct 14, 2017 at 9:22 AM, Scott Marlowe <scott.marlowe@xxxxxxxxx> wrote: > Please reply all to keep the list updated... > >> On Fri, Oct 13, 2017 at 9:07 PM, Scott Marlowe <scott.marlowe@xxxxxxxxx> >> wrote: >>> >>> On Fri, Oct 13, 2017 at 7:31 AM, Ashok Kumar Tiwari >>> <ashokt@xxxxxxxxxxxxxxxxxx> wrote: >>> > >>> > Hello team, >>> > >>> > My database goes in recovery mode every day three to four time , i am >>> > not able to find the reason, db basic configuration is max_connection = 1000 >>> > , db size 100GB and configure barman for backup. please help me. >>> >>> Check your server logs for the out of memory killer. With max conns of >>> 1000 it's quite possible your machine is running out of memory. >>> >>> Also please post any changed values in postgresql.conf >> >> >> >> > <ashokt@xxxxxxxxxxxxxxxxxx> wrote: >> >> i found in my db log file >> >> 2017-10-11 18:35:20.769 IST > LOG: checkpointer process (PID 31936) was >> terminated by signal 9: Killed >> < 2017-10-11 18:35:20.769 IST > LOG: terminating any other active server >> processes >> >> is this is out of memory? >> my server have 8 GB ram >> >> >> Very likely. Look around in the /var/log/messages around the same time for >> mention of the Out Of Memory killer (aka OOM). You're probably going to >> need to reduce max connections and then add something like pgbouncer or >> pgpool to reduce 1,000 outside connections down to something like 10 or 20 >> or so. Also look into to turning off the out of memory killer. It's a linux "feature" that is not desirable on a db server. Something like in this thread: https://ubuntuforums.org/showthread.php?t=2197016 -- Sent via pgsql-admin mailing list (pgsql-admin@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-admin
