I just deployed a new cluster of Openstack Victoria using PackStack on Centos Stream, same hardware but for some reason this new cluster runs terribly slow, from the dashboard to the actual instances (VM's) processing. Are there any possible causes I could be looking for or any logs I could provide that would provide some guidance?
Asked
Active
Viewed 199 times
0
-
In general: is the problem restricted to the (OpenStack) application layer, or is the problem also present on your base metal servers/compute nodes/hypervisors in the base OS? A fairly common cause for "slow" servers are DNS misconfigurations, missing forward and reverse DNS records, incorrect/absent (sub) domains etc. – Bob Apr 28 '21 at 13:04
-
Hey @Bob So it looks like it is only in the application layer or in the VM's. We have a docker build that takes about 15 minutes on our other Openstack Instance which is Rocky with Centos 7 on the same baremetal machines. But for some reason in this Victoria Centos Stream 12 baremetal node setup the builds are taking 45+ mintues. I did see some DNS mismatch records on our DC's so fixed them and also added the reverse lookup zone for our VM subnet in openstack. Going to run a couple tests and see if it helped at all. – schap Apr 29 '21 at 13:42