Hi,
I've been trying to use the FAS instance on publictest10, and it's acting very strangely. Basically, after a restart, it goes like this:
o FAS works fine, both from the web UI and using the API for a few minutes.
o After a few minutes, it starts to give 500 errors with stuff like this in the error_log:
[Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] File "/usr/lib/python2.4/site-packages/fas/model/fasmodel.py", line 6 1, in ? [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] PeopleTable = Table('people', metadata, autoload=True) [...] [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] File "/usr/lib/python2.4/site-packages/sqlalchemy/databases/postgres. py", line 462, in reflecttable [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] raise exceptions.NoSuchTableError(table.name) [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] NoSuchTableError: people
Very strange.
Any ideas what's going on here? If FAS on publictest10 is going through some development upheaval, I can set one up locally for development and try a public deployment of amber again in a few weeks.
Thanks,
-RN
On Mon, 7 Jul 2008 12:07:50 -0400 Robin Norwood rnorwood@redhat.com wrote:
Hi,
I've been trying to use the FAS instance on publictest10, and it's acting very strangely. Basically, after a restart, it goes like this:
o FAS works fine, both from the web UI and using the API for a few minutes.
o After a few minutes, it starts to give 500 errors with stuff like this in the error_log:
[Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] File "/usr/lib/python2.4/site-packages/fas/model/fasmodel.py", line 6 1, in ? [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] PeopleTable = Table('people', metadata, autoload=True) [...] [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] File "/usr/lib/python2.4/site-packages/sqlalchemy/databases/postgres. py", line 462, in reflecttable [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] raise exceptions.NoSuchTableError(table.name) [Mon Jul 07 08:23:07 2008] [error] [client 127.0.0.1] NoSuchTableError: people
Very strange.
Any ideas what's going on here? If FAS on publictest10 is going through some development upheaval, I can set one up locally for development and try a public deployment of amber again in a few weeks.
This still seems to be broken - anybody have any suggestions? I was hoping to let the websites team and others have a look at what I've got so far so I can get some feedback this week.
-RN
Ricky set up a FAS instance on pt9 that seems to work fine so far, so I'm done whining. :-)
Thanks!
-RN
On Tue, Jul 08, 2008 at 05:46:07PM -0400, Robin Norwood wrote:
Ricky set up a FAS instance on pt9 that seems to work fine so far, so I'm done whining. :-)
Hmmm, so what is the difference between the pt10 and pt9 deployments? We've been testing the bleeding-edge python-fedora package on pt10, so if that is causing the issues we definitely need to track it down asap.
luke
On 2008-07-08 08:48:57 PM, Luke Macken wrote:
Hmmm, so what is the difference between the pt10 and pt9 deployments? We've been testing the bleeding-edge python-fedora package on pt10, so if that is causing the issues we definitely need to track it down asap.
I manually updated python-fedora on pt9 to test that as well - so far, so good, although I still never figured out what was causing the NoSuchTable errors on publictest10.
Thanks, Ricky
On Tue, 8 Jul 2008 20:51:04 -0400 Ricky Zhou ricky@fedoraproject.org wrote:
On 2008-07-08 08:48:57 PM, Luke Macken wrote:
Hmmm, so what is the difference between the pt10 and pt9 deployments? We've been testing the bleeding-edge python-fedora package on pt10, so if that is causing the issues we definitely need to track it down asap.
I manually updated python-fedora on pt9 to test that as well - so far, so good, although I still never figured out what was causing the NoSuchTable errors on publictest10.
Well, this morning things started misbehaving on pt9, too...just differently. :-(
From web browser, if I go to:
http://publictest9.fedoraproject.org/accounts/
and log in, I get redirected to publictest10! If I browse back to pt9, it appears I did log in successfully. The equivalent thing happens if I log out from pt9.
Looking at the /etc/fas.cfg on pt9, this is probably the cause of that:
samadhi.baseurl = 'http://publictest10.fedoraproject.org'
I didn't fix it b/c I haven't familiarized myself with your puppet stuff yet.
Second, a few minutes ago I was getting 500 errors from that fas instance when logging in from my app, but now I can't reproduce it. :-/
Looking at the logs, it looks like the 500 errors are related to this error:
[Wed Jul 09 14:59:40 2008] [error] raise exceptions.DBAPIError.instance(statement, parameters, e, connection_invalidated= is_disconnect) [Wed Jul 09 14:59:40 2008] [error] ProgrammingError: (ProgrammingError) server closed the connection unexpectedly [Wed Jul 09 14:59:40 2008] [error] \tThis probably means the server terminated abnormally [Wed Jul 09 14:59:40 2008] [error] \tbefore or while processing the request. [Wed Jul 09 14:59:40 2008] [error] 'SELECT session.id AS session_id, session.data AS session_data, session.expiration_time A S session_expiration_time \nFROM session \nWHERE session.id = %(param_1)s' {'param_1': '620e1904fbcde17dacb96779e77ec9db223 48554'}
But, like I said, they aren't happening now. Curious. Maybe it just needed a second cup of coffee.
Thanks,
-RN
infrastructure@lists.fedoraproject.org