I was wondering to get answer of a confusion, let me portray a desing first:
I have a site with lots of servers and a storage, many of these servers are running production services. There are 10 server which are running a Virutal infastructure using vSphere 4.0 and using all enterprise features (So connected to SAN). We are going to build a DR site for this we have procure Recoverpoint.
As we know that recoverpoint will replicate the LUNs from Primary site to DR site. Now the question arise here that If Recoverpoint will replicate the LUNs from Primary site to DR Site and incase of Disaster we will failover to DR site, then why I need SRM for my VIrtual infrastructure to RUN from DR when Recovepoint will already replicate the LUNS to DR site and is continiously doing that.
- Please elaborate exactly where I need SRM ?
- What SRM is doing ?
- Can I replicate my VI to DR and can do the Failover incase of Disaster without SRM ?
Thanks in advance
- you need SRM to automate the DRP from VMware point of View (power on the VMs, manage the LUNs signatures, the order of boot for VMs, etc, etc)
- SRM is managing all before
- and the answer is yes, you can replicate your main site to drp site without the need of SRM, just take into consideration that all mentioned in answer 1 has to be done manually by someone., but you can actually do it.
Jose B Ruelas
Thanks for the reply; It means that I can run my services from DR site of a VI without SRM, the only thing is the manual internvention, nothing will be automated...right??
- Can you please tell me exactly about what processes or procedures I'll be doing manually in case of Failover ?
one more question:
- Due to hardware resources shortage, I am also considering to virtualize few services on DR site which are hosted as Physical in Primary site. I believe I can do this also without SRM. What processes I need to do manually?
In short, the manual steps you will need to do: shut down any vms that might be running prior to bring up the productions vms, disable SAN replication, makeactive the replicated LUNs, bring up the VMs and re-ip address if necessary. This is all automated by SRM.
There were a couple of things that was left out that your can do with SRM: you can up different recovery plans you do not have to create an all or nothing DR recovery plan, you can actually use the DR site for non-critical VM machines such as development because they will be shut down automatically and what I consider the most importnat feature you will be able to test your DR plans with impacting production.
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
SRM is a DR automation tool. It doesn't do replication but it does interface with replication tools like RecoverPoint however.
Here is what your DR workflow would look like without SRM while using RecoverPoint for an actual failover:
1. Tell recoverpoint to failover
2. Ensure LUNs are presented to your ESX hosts
3. Tell ESX hosts to mount/resignature the LUNs presented
4. Register all of the virtual machines stored on the LUNs (this can take a while or be really quick if you're good with scripting)
5. Power up each VM based on your priority
6. As each VM is powered up you may need to change the IP address and DNS entries to be appropriate for the DR site
7. You may need to also change/adjust portgroup settings
Testing your DR plan will be a bit more challenging. You'll need to place your RPA into logged image access mode, present said image to your ESX servers; mount; follow the above workflow to the letter. Once your test cycle is complete you'll need to unregister all of the VMs and put the RPA back into normal mode (forget the name off the top of my head.)
What SRM will do is automate every one of these tasks for you. It will interface with your RPA at either site depending on what you're doing; it will register your VMs and potentially change IP addresses. It will let you establish priorities and execute arbitrary scripts (for example sending a power-on message to a virtual machine that is physical in your primary site.) It also provides a nice and easy test facility that lets you do DR testing pretty much at will.