somejfn commented on issue #2984: kvm: reset KVM host on heartbeat failure URL: https://github.com/apache/cloudstack/pull/2984#issuecomment-434654783 I guess VM-HA is handled differently on managed storage right ? I mean Ceph has image locking builtin and I hope Solidfire can do the same to guarantee exclusive access. On NFS the risk to keep HA is greater than its benefits so sadly I'll have to disable it and humans handle host failures On Tue, Oct 30, 2018 at 9:23 AM Wido den Hollander <notificati...@github.com> wrote: > I understand @somejfn <https://github.com/somejfn> , but 22 minutes is a > very long time. In this case you can't blame CloudStack for the fact that > it would reboot, > > Merely stopping the Agent is no guarantee that the VMs are really down. > > You can argue on how long the timeout should be, 5, 10, 15 minutes, but at > some point a HV would need to kill itself so that we can be sure that VMs > are dead. > > NFS does not provide proper locking of images, so two VMs could run from > the same disk. > > Downtime is not nice, but data corruption is horrible. > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <https://github.com/apache/cloudstack/pull/2984#issuecomment-434297763>, > or mute the thread > <https://github.com/notifications/unsubscribe-auth/AOflqNZ8__gqMF_oJC36HMKTLH6cUkHCks5uqFK6gaJpZM4YBbJC> > . >
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
