GitHub user cookesan added a comment to the discussion: Is secondary management 
server able to handle infra in data center 1

I would not model these as primary and secondary management servers. In 
CloudStack they are management-server nodes sharing the CloudStack database and 
agent connectivity.

If only the management server in data center 1 is down, another management 
server can take over as long as it can reach the same CloudStack database and 
the data-center-1 hosts, system VMs, and storage management endpoints. The 
agent code is built around a list of management server addresses and can 
migrate an agent connection to another reachable management server; the 
database config is common CloudStack state, not per-zone state.

If the whole data center 1 is unreachable, a management server in data center 2 
will not be able to provision resources in data center 1 just because it is 
still running. It still needs network reachability to the hosts and the 
management/storage networks for that zone.

So they do not need to be on the same L2 network, but they do need routable 
connectivity for the CloudStack control plane: shared database access, 
API/UI/load-balancer access as you design it, and 
management-server-to-agent/system-VM reachability for the zones they are 
expected to control.


GitHub link: 
https://github.com/apache/cloudstack/discussions/13500#discussioncomment-17455833

----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]

Reply via email to