VMware vCenter Site Recovery Manager (SRM) does many things well when it comes to automating the recovery process for virtualized environments that simultaneously eliminates the complexity of creating and testing recovery plans. But what SRM does not do is provide control of and visibility into the applications within VMs at the recovery site. That changes today as organizations now have a new option to enhance their SRM deployment with the latest release of Symantec ApplicationHA.
Those familiar with VMware vCenter SRM may already be aware and familiar with how SRM handles the recovery process for VMs. Should a failure or disruptive event occur at the primary site (site outages, storage array needs to be taken offline for maintenance, etc.) an administrator can initiate a failover to the secondary site of all of the VMs that are within an SRM Protection Group.
But when an administrator initiates a failover to the recovery site of that SRM Protection Group, SRM is not natively aware of the status of the applications within any of the VMs it just failed over to the recovery site. SRM assumes that the applications hosted by the VMs will come up normally and resume processing without any issues.
This assumption may be appropriate when failing over VMs that function as file, print or web servers. But as more organizations virtualize business critical applications like Microsoft Exchange, SharePoint, SQL Server and others, it is not a given that these applications will automatically restart correctly at the recovery site. If anything, this is a dangerous assumption to make especially if they just had their service interrupted by a server failure or storage outage at the primary site.
So if using SRM to failover these applications to a recovery site, organizations should assume the following are NOT TRUE:
- If the VM is up then the application is also up
- If the application processes are up then the applications are functional
- If the VM went through the proper start-up sequence then the application started correctly
Aggravating the problem, whether one takes a cursory glance or a detailed examination of the vCenter SRM console, there is almost no information as to the status of the applications on individual VMs. So if the VMware administrator sees that all of the VMs are up on the vCenter console, he could prematurely and incorrectly conclude that all of the applications hosted by these VMs are also up and operational.
It is this lack of control and visibility that today’s release of Symantec ApplicationHA 5.1 SP2 addresses. ApplicationHA already provided control of and visibility into applications hosted on VMs so that administrators could check the status of applications, verify they were operational and even start or re-start processes on them. However should the application be on a VM that has failed over to another site courtesy of SRM, ApplicationHA lost visibility into the application’s status and the ability to execute applications commands.
That loss of control and visibility into VMs that move with SRM failovers ends with this release of ApplicationHA. Now should an application that is hosted on a VM failover to the recovery site, ApplicationHA’s visibility and control go with it. By going to the ApplicationHA tab in vCenter after the host VM boots up, the administrators may check the status of the application, verify it is operational, that application start-up processes occurred in the correct order and re-start any processes that did not come up when the VM restarted.
SRM is a powerful tool within the VMware suite of virtualization software but organizations need to be aware of what functionality it can and cannot provide as they start to virtualize business critical applications and look to use SRM to recover them. Primarily it is important to note that SRM’s primary objectives are to ensure that an application’s data is ready to recover and that the VM hosting the application has successfully re-started.
By itself SRM does not have native control or visibility into the status of any of the applications on any of the VMs. ApplicationHA now gives administrators that additional control and visibility they need to have as they virtualize their business-critical applications regardless of whether these applications reside at the primary or recovery site.