internal server error (timeout) while creating an OpenVZ container on a slave

[pille]

when i try to create a new OpenVZ container from a template and chose it go on a slave cluster node, the ajax request done at this point fails with an Internal Server Error 500:

[4414]ERR: 24: Error in Perl code: 500 read timeout

however - i can always create it on the master cluster node and migrate it afterwards to the slave node.

 

[dietmar]

Please check that 'pvetunnel' and 'pvedaemon' services are running on all nodes.

 

[pille]

yes, they are up and runnig on all nodes. i guess it's one of those unreproduceable problems.
in the meantime after trying it again and again suddenly it worked.
perhaps there's network congestion due to cluster sync?!
what's the timeout value the error message talks about?

 

[dietmar]

yes, they are up and runnig on all nodes. i guess it's one of those unreproduceable problems.
in the meantime after trying it again and again suddenly it worked.
perhaps there's network congestion due to cluster sync?!
what's the timeout value the error message talks about?

AFAIK its 10 seconds.

 

[pille]

cannot find it in the code after some quick searches, but 10secs seems reasonable when i'm counting.

where's it in the code? i want to increase it a bit to check if that helps for all cases.

 

[dietmar]

where's it in the code? i want to increase it a bit to check if that helps for all cases.

/usr/share/perl5/PVE/ConfigServer.pm

    # set longet timeout for local connection
    my $timeout = $cid ? 10 : 120;