Hi Bryan, what really will help if you can provide the management-server logfile when something like this happens. Did something changed on your setup on Monday?
How did you implement HA with Linstor primary storage? regards, Swen -----Ursprüngliche Nachricht----- Von: Bryan Tiang <[email protected]> Gesendet: Freitag, 26. Januar 2024 06:51 An: Vivek Kumar via users <[email protected]> Betreff: URGENT: Unstable VM and VR Performance with Cloudstak and Cant seem to find root cause Hi Community, Urgently need help on this. We are experience unstable performance with Cloudstack and have been having this issue since Monday... We're facing this error frequently and randomly. Unable to get answer that is of class com.cloud.agent.api.StartAnswer We encounter this during the following scenarios: # When VM failover to another host and is not able to start # When creating and starting a new VM # When starting an existing VM from Stopped state # When starting a stopped Virtual Router # When starting a new Virtual Router It happens very randomly and we can’t seem to identify a pattern. For example, when creating a VM fails, we literally just spam repeat the process and suddenly it will work. Or, sometimes we restart the VPC with Virtual Router Cleanup and it will suddenly work again. What we've done #Restarted management server #Removed cloudstack-agent and its directories in all Hypervisors #Increased CPU and memory for Virtual Router offering #Restarted Linstor Storage and Sattelite We are using Cloudstack 4.18.1 + Linstor + Ubuntu. On the host we applied CIS Benchmark hardening for Ubuntu 22.04 and AMD Memory Guard enabled. Regards, Bryan
