> As far as i know, this problem is caused by having sqlalchemy.pool_recycle
> value larger than my.cnf [mysqld] connect_timeout
I wasn't able to verify this in my environment.  I set
sqlalchemy.pool_recycle to 30 and connect_timeout to 60 in my.cnf and
I still was getting 'MySQL server has gone away' errors. I tested this
by setting my.ini and sqlalchemy variables then restarting MySQL and
Pylons.  After making an initial page request, I waited waited for set
periods before making a second (non-cached) request. In each test, the
errors started appearing right after the wait_timeout. After trying
various configurations of wait_timeout, connect_timeout, and
sqlalchemy.pool_recycle, the only variable that appeared to have a
direct impact on my particular manifestation of this problem was
wait_timeout. This makes some sense after reading the MySQL docs on
this error: http://dev.mysql.com/doc/refman/5.0/en/gone-away.html

-Chris

On Jan 21, 9:21 pm, Didip Kerabat <[email protected]> wrote:
> As far as i know, this problem is caused by having sqlalchemy.pool_recycle
> value larger than my.cnf [mysqld] connect_timeout
>
> - Didip -
>
> On Thu, Jan 21, 2010 at 8:15 AM, Ross Vandegrift <[email protected]> wrote:
> > On Thu, Jan 21, 2010 at 12:05:43AM -0800, grassoalvaro wrote:
> > > mysql.config
>
> > > wait_timeout=90
> > > max_allowed_packet=1M
> > > connect_timeout=300
>
> > > I fixed my pylons configurations set pool_size to 60 but didn't help.
> > > Application is very simple, just 2 controllers, nothing complicated
> > > (this app .I have many others with the same problem. Currently i'm
> > > using cron to sending one GET request on every minute but this is
> > > ugly).
>
> > In cases where I've run into this, I've turned the pool_recycle time
> > way down - mine is set to 30, and this fixed the issue.
>
> > Note that this is a workaround for another problem - in my case, I
> > happen to know that there is a firewall that prevents long-lived TCP
> > sessions from sitting idle.  This causes pooled connections to "go
> > away" without notifying the application stack.
>
> > In theory, the better solution would be to fix that firewall's
> > config.  In practice, reducing the pool_recycle time doesn't affect my
> > very light database load.
>
> > Ross
>
> > --
> > Ross Vandegrift
> > [email protected]
>
> > "If the fight gets hot, the songs get hotter.  If the going gets tough,
> > the songs get tougher."
> >        --Woody Guthrie
>
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.9 (GNU/Linux)
>
> > iEYEARECAAYFAktYfYcACgkQMlMoONfO+HAA/wCeI3s0bCFHkHB9GvQOdF3NThUO
> > 5+oAoM3oPevLy4hENZF3ny4jF0MhF/3Q
> > =dq1w
> > -----END PGP SIGNATURE-----
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"pylons-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/pylons-discuss?hl=en.

Reply via email to