Closed kustodian closed 9 years ago
For versions <9.2, that do not support DROP INDEX CONCURRENTLY
, to
soften locks impact, pgcompact sets statement timeout locally for its
transaction to 500ms and performs several retries. The first two
errors show that.
However, the third error tells that pgcompact was cancelled by the
user on the client side. Probably by Ctrl+C
.
2015-03-12 00:24:02 CET [22772]: [208-1] ERROR: canceling statement
due to user request
The thing is pgcompact is run via cron and no one was logged in at that time on the server to send a kill signal.
On Mon, Mar 30, 2015, 23:25 Sergey Konoplev notifications@github.com wrote:
Closed #6 https://github.com/grayhemp/pgtoolkit/issues/6.
— Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#event-269114460.
I might be missing something, but that is what I see is happening.
On Tue, Mar 31, 2015 at 4:33 AM, Strahinja Kustudic < notifications@github.com> wrote:
The thing is pgcompact is run via cron and no one was logged in at that time on the server to send a kill signal.
On Mon, Mar 30, 2015, 23:25 Sergey Konoplev notifications@github.com wrote:
Closed #6 https://github.com/grayhemp/pgtoolkit/issues/6.
— Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#event-269114460.
— Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#issuecomment-88054094.
Kind regards, Sergey Konoplev PostgreSQL Consultant and DBA
http://www.linkedin.com/in/grayhemp +1 (415) 867-9984, +7 (499) 346-7196, +7 (988) 888-1979 gray.ru@gmail.com
When this had happened I analyzed everything and couldn't find anything which would send SIGINT to pgcompact, but I might have missed something. Leave it closed, if it happens again I'll check once again, and report back if it's pgcompact related.
On Tue, Mar 31, 2015, 11:54 Sergey Konoplev notifications@github.com wrote:
I might be missing something, but that is what I see is happening.
On Tue, Mar 31, 2015 at 4:33 AM, Strahinja Kustudic < notifications@github.com> wrote:
The thing is pgcompact is run via cron and no one was logged in at that time on the server to send a kill signal.
On Mon, Mar 30, 2015, 23:25 Sergey Konoplev notifications@github.com wrote:
Closed #6 https://github.com/grayhemp/pgtoolkit/issues/6.
— Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#event-269114460.
— Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#issuecomment-88054094.
Kind regards, Sergey Konoplev PostgreSQL Consultant and DBA
http://www.linkedin.com/in/grayhemp +1 (415) 867-9984, +7 (499) 346-7196, +7 (988) 888-1979 gray.ru@gmail.com
—
Reply to this email directly or view it on GitHub https://github.com/grayhemp/pgtoolkit/issues/6#issuecomment-88118895.
Today on one of our Postgres 9.1 database pgcompact (1.0.2) exited with this error:
Here is what was logged in the postgres log:
As you can see a temp index was created, but pgcompact due to a time out couldn't drop the old index (and after the second retry it cancelled the DROP query). The problem is that pgcompact exited after this and it didn't clean up the temp index.
Is this normal? Shouldn't pgcompact continue working even if one index fails? I guess pgcompact should at least drop the temp index?