Re: log_destination reload/restart doesn't stop file creation
От | Amit Kapila |
---|---|
Тема | Re: log_destination reload/restart doesn't stop file creation |
Дата | |
Msg-id | CAA4eK1L95f5F6w1Nc-P3_iww7rEst4ymJNX4XddKt8oPYKCB5Q@mail.gmail.com обсуждение исходный текст |
Ответ на | log_destination reload/restart doesn't stop file creation (Thom Brown <thom@linux.com>) |
Список | pgsql-bugs |
On Tue, Aug 14, 2018 at 6:03 PM Thom Brown <thom@linux.com> wrote: > > Hi, > > I've tested the following on git head. When changing log_destination > and reloading, the old destination file continues to be made, just not > populated with anything. That means at every file rotation, 2 files > are created. > > For example: > > log_destination = 'stderr' > > I get this in my log directory: > > postgresql-2018-08-14_131640.log > > If I change it to csvlog and reload, I end up with: > > 0 -rw------- 1 thom thom 0 Aug 14 13:19 postgresql-2018-08-14_131900.log > 4 -rw------- 1 thom thom 194 Aug 14 13:19 postgresql-2018-08-14_131900.csv > > So I get the csv file, but it's still producing the .log file which > remains 0 bytes. The same happens in reverse. (i.e. I end up with an > empty .csv file and a populated .log file). > I don't see this problem happening when you are changing log_destination from csvlog to stderr and reload the conf via pg_reload_conf(). The reason why it is happening when log_destination is set to csvlog is that there are some cases as explained in Kuntal's email when we can only write in the log file. So, I don't see any bug here, but I might be missing something here. -- With Regards, Amit Kapila. EnterpriseDB: http://www.enterprisedb.com
В списке pgsql-bugs по дате отправления: