This is the first post for my “Implementing VMware Site Recovery Manager with EMC VMAX 20K storage replication” series.
My focus at the start of deploying SRM with VMAX 20K was understanding the environment and what I wanted to achieve. I did this by understanding the layout of the environment and what was required. For me, simply put, it looked like this:
Once I was confident in this aspect, I had some reading to do – having never deployed this particular configuration, using SRDF replication! For this particular task, there was only one man to read up on @CodyHosterman – he has been there, done it, got merchandising rights and wrote the book(s)! This blog of his was particularly useful!
After much reading and re-reading (I’m not a VMAX expert so parts were hard to understand, but I got there…) the key areas for me were:
– Get the software with correct versions supported for my environment.
– Get two sites configured with SRM components.
– Must have asynchronous array replication working (SRDF/A)
– Must utilise remote Solutions Enabler.
Here is a list of software and versions I used:
- VMware vCenter Server 5.5 u1c (Limitation of environment due to Vblock Release Candidate Matrix level).
- VMware SRM 5.5.1 (5.8 isn’t supported on u1c)
- Microsoft SQL Server 2012
- Windows Server 2012 R2
- EMC Solutions Enabler 8.0.1 (For Array replication)
- VMware Storage Replication Adapter 5.6
- EMC Virtual Storage Integrator 5.7.1
The implementation I decided on, derived from Codys excellent guide, is as follows:
1) Install SQL Servers for SRM.
2) Install SRM on both sites.
3) Configure SSL certificates for SRM
4) Link SRM Sites
5) Install Solutions Enabler software on Array Managers and SRM Servers.
6) Install and configure VMware Storage Replication Adapters for EMC SRDF.
7) Install EMC Virtual Storage Integrator
8) Enable and create VMAX SRDF/A replicated volumes
9) Configure Datastores for SRM & Resource Mappings
10) Create Protection/Recovery Groups
11) Perform a Recovery
I’m going to write the remaining parts of this guide for the above sections to show an end to end example implementation; with the end goal being to run some test fail overs to prove confidence in the underpinning virtual disaster recovery infrastructure.