Custom Query (87 matches)

LabWiki? on ​http://emmy9.casa.umass.edu:4000 response slows down with more than 10 experiments running.

top returns:

top - 11:46:51 up 6 days, 18:31,  3 users,  load average: 9.31, 7.79, 6.78
Tasks: 613 total,   3 running, 609 sleeping,   0 stopped,   1 zombie
Cpu(s):  4.2%us,  2.0%sy,  0.0%ni, 86.5%id,  7.2%wa,  0.0%hi,  0.2%si,  0.0%st
Mem:  12291136k total, 11220328k used,  1070808k free,   644620k buffers
Swap: 12569596k total,  1234756k used, 11334840k free,  1227576k cached

The response is considerably slow when many experiments are running.

Is there anything specific we should me monitoring or logging when the socket timeout occurs?

This is a ticket to shadow the "FiberError?" Labwiki crashes tracked in ​http://mytestbed.net/issues/1626.

Experiment controller is not running experiments on nodes, issue seen around the morning of Oct 2nd, around the same time as Openfire was restarted. Experiments can be run on new slices but not on the same slice more than once. Restarting of RCs does not work either. Deleting and adding resources back to the slice still does not solve the problem.