Workflow: prepare for an NSP DR switchover
Description
Before you perform an NSP DR switchover, you must ensure that the conditions are in place for a successful reversal of the primary and standby NSP cluster roles.
The following sequence of high-level actions describes how to ensure that the primary and standby NSP clusters are in the correct state for a successful DR switchover.
Stages
| |
|
1 |
Use the NSP UI to verify that there are no outstanding Critical or Major alarms against any NSP cluster.
|
2 |
Verify that each pod listed in the following table is in the state shown.
|
Pod |
Primary cluster |
Standby cluster |
|
nspos-zookeeper |
Running |
Running |
|
nspos-postgresql
1 |
Running |
Running |
|
nspos-tomcat |
Running |
Running |
|
nspos-prometheus |
Running |
Running |
|
nspos-kafka |
Running |
n/a |
|
nsp-tomcat |
Running |
Running |
|
nspos-keycloak |
Running |
Running |
|
nspos-oauth2-proxy |
Running |
Running |
|
nspos-asm-app |
Running |
Running |
|
nsp-role-mgr |
Running |
Running |
|
nsp-file-service-app
2 |
Running |
Running |
Notes:
Restarts during switchover to change role and functions If present; pod not required in all deployment types
|
3 |
Perform
How do I identify the NSP cluster DR roles? on each cluster; use the statusAll command option to ensure that each cluster has the correct role, which means that one cluster has the primary role, and the other the standby.
|
4 |
Perform
How do I check NSP database synchronization? to ensure that the primary and standby databases are 100% synchronized.
|
5 | Verify that the standby nsp-file-service-app pod can SSH to the primary nspos pod.
-
Enter the following as the root user on the NSP cluster host in the standby data center:
# kubectl exec -n $(kubectl get pods -A | awk '/nsp-file-service-app/ {print $1;exit}') nsp-file-service-app-0 -it -- bash ↵
A root shell opens on the nsp-file-service-app pod.
-
Enter the following to attempt an SSH connection to the primary nspos pod:
# ssh -i /opt/nsp/apps/nsp-file-service-app/nspSSHcerts/hostcert -p 8566 -o StrictHostKeyChecking=no address ↵
where address is the NSP cluster advertised address
If the command returns an error message, the SSH connection attempt has failed.
-
If the connection attempt fails, resolve the connection issue; you must not attempt a DR switchover unless the problem is resolved.
|
6 |
Close the open console windows. |