Scenario:
Recovery Plan 1: APP1, DB1, WEB1
Recovery Plan 2: APP2, DB2, WEB2
SRM can do datacenter migration/disaster avoidance:
1. If failover is initiated, PROD VMs do a last replication, power down PROD VMs and power on DR VMs?
2. If Recovery Plan 1 is initiated failover, VMs in the Recovery Plan 2 will not be affected right. same if recovery Plan 2 is initiated?
3. Will SRM stop the replication of Recovery Plan 1 VMs (vSphere Replication/Array Based Replication) only and not the Recovery Plan 2 VMs (Recovery Plan 2 VMs replication ongoing)?
4. If Recovery Plan 1 VMs are powered on the DR site and reprotect is initiated for this recovery, only Recovery Plan VMs 1 will be replicating now from DR to PROD without affecting the PROD to DR replication of Recovery Plan 2 VMs?
Hi
1. If failover is initiated, PROD VMs do a last replication, power down PROD VMs and power on DR VMs? (Yes it will try to replicate the changes if the prod storage is available)
2. If Recovery Plan 1 is initiated failover, VMs in the Recovery Plan 2 will not be affected right. same if recovery Plan 2 is initiated? ( As long these VMs are on separate Data Stores will not effect each others)
3. Will SRM stop the replication of Recovery Plan 1 VMs (vSphere Replication/Array Based Replication) only and not the Recovery Plan 2 VMs (Recovery Plan 2 VMs replication ongoing)? (Yes, Once you done the failover you can reprotect the VMs so it will do the reverse replication)
4. If Recovery Plan 1 VMs are powered on the DR site and reprotect is initiated for this recovery, only Recovery Plan VMs 1 will be replicating now from DR to PROD without affecting the PROD to DR replication of Recovery Plan 2 VMs? (Yes)
Regards
Umar
2. If Recovery Plan 1 is initiated failover, VMs in the Recovery Plan 2 will not be affected right. same if recovery Plan 2 is initiated? ( As long these VMs are on separate Data Stores will not effect each others)
>> separate datastores in PROD or DR site?
current setup is all VMs are residing in PROD VSAN while replicated DR VMs are in DR_VSAN.
So If my scenario, I triggered Recovery Plan 1; Recovery Plan 2 VMs will also be affected since they are residing in 1 datastore?
Current replication is through vSphere Replication but eventually some of will be using array replication.
Please confirm if these will be the same regarding of replication approach
As long as what you have listed are separate Protection Groups (Web1, Web2, App1, App2, etc) then regardless of replication type they will not impact each other.
Right now they all reside on the same datastore, but replication is at the VM level so that doesn't matter. If you were using array based replication they couldn't all be on the same datastore and belong to separate protection groups.
See this post for PG design/organization questions/answers SRM Protection Group Design Considerations - VMware vSphere Blog
As long as what you have listed are separate Protection Groups (Web1, Web2, App1, App2, etc) then regardless of replication type they will not impact each other.
Right now they all reside on the same datastore, but replication is at the VM level so that doesn't matter. If you were using array based replication they couldn't all be on the same datastore and belong to separate protection groups.
See this post for PG design/organization questions/answers SRM Protection Group Design Considerations - VMware vSphere Blog
>> so it means even they are hosted in the same datastore, as long as they are grouped in separate protection groups; If I triggered a recovery/failover on Recovery Plan Group 1, only VMs in that group will power off to the DR site and and VMs in Recovery Plan 2 will not be affected?
That is Correct ChevUribeChevUribe.
The statement you have put there is correct.
so it means even they are hosted in the same datastore, as long as they are grouped in separate protection groups; If I triggered a recovery/failover on Recovery Plan Group 1, only VMs in that group will power off to the DR site and and VMs in Recovery Plan 2 will not be affected?
Just to avoid any confusions, this applies to vSphere replication only used along with SRM.
When we use Array based replication and SRM the situation is different.