version: 8.3
The other day, my DB stop processing request. It still accepts connections but not processing those.
So I quit all client connections from client yet those process still alive on SQL Server.
I tried to Stop DB by issuing pt_ctl STOP -m fast but failed to shut down database.
Next I issued pg_ctl stop -m immediate. I think this one killed all processes that were from client.
The returning message says the server was stopped.
So I issued ps -ef | grep postgres but master and 4 child processes were still there...
502 150 1 0 0:10.70 ?? 0:12.69 /Library/PostgresPlus/8.3/bin/postgres -D /data
502 165 150 0 0:03.71 ?? 0:05.15 postgres: logger process
502 171 150 0 0:44.15 ?? 1:06.28 postgres: writer process
502 72026 150 0 0:00.15 ?? 0:00.87 postgres: postgres mydata [local] VACUUM
502 81413 150 0 0:00.48 ?? 0:06.52 postgres: postgres mydata 127.0.0.1(56760) COPY
I then tried to kill the following process
502 81413 150 0 0:00.48 ?? 0:06.52 postgres: postgres mydata 127.0.0.1(56760) COPY
even with -s 15 I could not kill....
So what I end up doing was to hard reboot the machine...
I have questions:
1) Did I have better choice than hard-reboot the machine? (soft reboot didn't work).
2) The COPY command I was tying to kill is local request. Is this the process postgres trying to write changes to the disk? If so, what could I have done to data?
Thank you very much for your time.