Mid tier Active Passive Cluster

[raj231]

Hi, If in case services on active node are down but the SAS IDP  is being supported by passive node, then to bring the services up on active node do we need to restart services on both active and passive nodes? If this is the configuration of mid tier currently being used, Please advise

[JuanS_OCS]

Hello @raj231,

I am not sure of your question.

If you ask if you need to start 2 nodes of your cluster, to start providing service: not you don't. Just start the services on the right order in one node, and your SAS environment will provide services (if your load balancer works properly).

As soon as SASServer1_1 (and better SASServer1_2 as well) is started in one node, you can start the rest of services in the second node.

[raj231]

Apologies, little bit confused here. The configuration is active passive SAS mid tier with two nodes one as primary and the other secondary. The primary nodes services are down but the SAS Information Delivery Portal is now running without any down time as the secondary node is supporting it which became active now.

Now, can you please suggest in order to bring the services up on primary node, do I need to bring down the services of the secondary fail over node as well and then restart the services in the order of Primary and then Secondary, please advise, thanks

[JuanS_OCS]

Hello @raj231,

thank you, that helps a bit to clarify the question.

I still miss details on how the active-passive cluster is configured: load-balancer configuration, hostname and network card configurations, etc.

Depending on those configuration, perhaps it is possible to start the primary node, normally, and switch down the secondary node. Or perhaps not at all and part of your configuration could become corrupted.

My current understanding is that perhaps you do not have this in-depth knowledge, therefore, perhaps your best option is to learn by experience, meaning:

• Notify your users that SAS environment will have maintenance and there might be some downtime, or even provide a more disclosed explanation.
• Make a snapshot for  your virtual machines and/or backup of your SAS deployment (metadata, SAS Content Server, WIP db..)
• Start the primary node
• Stop the secondary node
• Validate. If something does not work, perhaps you might need to recover the environment.

If you want to go through the safe mode, then just plan downtime, stop the secondary node, and start the first node.