handle restarts of DBMS gracefully
Brought to you by:
sits
I'm using CodeStriker with PostgreSQL and have a
neasily reproducible problem:
1. access CS topic
2. do /etc/init.d/postgresql restart
3. try to access CS again: you get a "failed to connect
to db" error
It's a nasty problem because whenever (2) happens (for
unrelated but unavoidable reasons), the only solution
is to also restart apache.
I think the reason for this bug is that either CS or
DBI layer caches connections but doesn't realize that
connection has become invalid because the server has
been restarted. It would be great if it could try to
reconnect to the DB server anew instead of just giving
an error.
Thanks in advance!
Logged In: YES
user_id=208928
Making this a feature request, as strictly speaking, this is
not a bug.