Description of problem: Working on small scale ( 1 controller ,1 networker and 2 computes ) and it seems 1. ML2 +L2 pop i can only reach about 60 VMs (cirrus/nano), any known bug ? 2. ML2 +L2 pop disabled reach 120 VMs ( with/without multi neutron workers ) With L2Population disabled I was able to achieve more guest launches in parallel. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1.See above 2. 3. Actual results: Expected results: Additional info:
Ofer, any details on what issues you saw when launching lots of instances with l2pop enabled? As per Joe findings, there is some bug due to SIGTERM sent to OVS agent as in: https://bugzilla.redhat.com/show_bug.cgi?id=1136969 Any other specific issues you have in mind?
OK, I assume the bug is a duplicate and will be fixed in the bext async release that will receive upstream rebase that includes the fix. *** This bug has been marked as a duplicate of bug 1136969 ***