While designing the SRM environment, I encountered a confusion. What is the difference between Priority groups and dependencies in SRM. As I understand, priority groups are definedto segregate each set of VMs to be Powered OFF/Powered ON sequence during Disaster Recovery and dependencies are defined to Power OFF/Power ON the VMs in the sequence if there is any dependencies between the applications/VMs. My question. What is the difference between putting 5 VMs in 5 different priority groups and putting 5 VMs in same priority group & defining the sequence of Power OFF/Power ON in the way we wanted?
Priority groups determine the order in which the VMs will be bought up.
VMs in priority group is bought up first and then in P2 and so on.
Now, in P1 group there might be a app server and DB server and application is dependent on DB server. This is where dependencies come in.
So if you set App Server is dependent on DB server, then we make sure that DB server is powered up and available first post which app server is powered on.
Hope that clarifies it.
Thanks for the reply. I do know, how the priority groups works and what happens when we opt in dependencies. But my question is, what is the difference,if we have 5 VMs and putting those 5 VMs in different priority groups serves the purpose of putting 5 VMs in single priority groups and defining dependencies. So if we are putting 5 VMs in 5 different priority groups, VMs will Power ON as priority 1 > 2 >3 > 4 > 5. But if we put dependencies, this can be achieved by putting 5 VMs in single priority group. I really didn't understand the difference between these 2 options.
Two ways to do the same thing. There is nothing to differentiate here. The options of priority and dependencies were provided for the reason in my last comment.
If you want to put them in same Priority group and define dependencies rather than splitting them up, you can go ahead.
If you want to split them, you can still do it.
There is absolutely no point differentiating this use case.
I would say there is a slight difference in priority group and dependencies
Priority group is up-to a point where it validates poweron and shutdown order . So once SRM observes the VM is powered on, it proceeds further without verifying if OS and all the required services are up and running.
When you set a dependencies you verify if the required services dependent on another server is started.
"If a virtual machine depends on services that run on another virtual machine in the same protection group, you can configure a dependency between the virtual machines."
I think they're supposed to work together in order to scale.