Great pointer, Daniel, thank you!
---
Nux
www.nux.ro
On 2022-12-05 16:26, Daniel Augusto Veronezi Salvador wrote:
Hello, Ranjit and Nux
About the second point in the last message: there is no need to edit
the file 'kvmheartbeat.sh' to avoid restarting the host. PR #4586[¹],
released in 4.16.0, introduced the property
'reboot.host.and.alert.management.on.heartbeat.timeout' in
'agent.properties'; if true (default) it will reboot the host if
something goes wrong with the heartbeat; on the other hand, if set to
false, it will not reboot the host. After changing the properties, it
is necessary to restart the agent service.
Best regards,
Daniel Salvador (gutoveronezi)
[¹]https://github.com/apache/cloudstack/pull/4586
<https://github.com/apache/cloudstack/pull/4586>
On 05/12/2022 12:38, Nux wrote:
Hello,
For HA to work you need to:
1 - add a service offering with "HA" box enabled
2 - you need to have some NFS storage for the heartbeat/fencing
mechanism - you don't need to use it for VMs, but it needs to be
present and it needs to be super reliable or hypervisors might reboot
if they see it goes away - this behaviour can be customised if you
want to (and know what you are doing):
https://github.com/apache/cloudstack/blob/a63b2aba7aa3948f78d280a356c550c6638a137b/scripts/vm/hypervisor/kvm/kvmheartbeat.sh#L162
HTH
---
Nux
www.nux.ro
On 2022-12-02 07:02, Ranjit Jadhav wrote:
Hello Guys,
We are also struggling with this HA thing any guidance will be highly
appreciated.
Regards,
Ranjit
On Thu, 1 Dec 2022 at 20:50, pspa...@hotmail.com
<pspa...@hotmail.com>
wrote:
Hi,
I have build new infra with Ceph storage everything working well.
But VMs HA not working. Can anyone guide me.
Regards Pradeep Kumar
