I was told that the VNX 5200 can replicate to an NX4. We have two NX4's, one in Production and one at a DR site. We use the SRA's for VMWare Site Recovery Manager. We are all NFS for our VMWare to NX4 over gig Ethernet.
I am going to be migrating production to a VNX5200 NFS over 10gig Ethernet for VMWare. Though the VNX5200 has a lot more storage than the NX4. My question is if we use replication, does the other side have to have the same exact filesystem size, even if the filesystem is not full? I am looking at one of my VNX5200 filesystems to be 16TB but the other end only has 12TB max. But if I only have 4TB out of 16TB used for example, would that be ok?
Otherwise I would have to break things up and I really don't like to do that because its harder to manage. I also want to use all 16 disks in each tier on the VNX5200 in RAID to take advantage of more spindles for higher IOPS, rather than split each tier into two 8 disk raids.
Thanks for the input
Oh we will use storage vmotion to move VM's from the NX4 to the VNX5200. However that covers our Production site. The DR site we did not invest in a VNX... it still has the NX4 and still will have the NX4.
So when the VM is moved off our existing production NX4, it will no longer be replicated to our DR NX4. So I need to create another new replication session and a new file system at the DR site to replicate from VNX to NX4.
Eventually the NX4 at the Production site will only be used for things like test machines or VEEAM backup images.
Maybe NFS replication isn't the best answer. Maybe I need to migrate to the VMWare replication virtual appliance. Then I could upgrade to vSphere 5.5. I'm stuck at 5.0 because there's no 5.5 SRA's for the DR site's NX4 at this time.
I can move a DAE from our Prod site to DR site eventually, but of course a certain percentage of storage has to be migrated to VNX first. Additionally starting replication over from scratch over the 40mb WAN will take weeks to move a few TB's plus about 135GB change rate per business day.
start with smaller file system on VNX5200, say 10TB (leave some room for internal snapshots). As you free up space on production NX4, move DAEs to DR NX4 (work with your account team so not to invalidate your support contract) and then increase VNX5200 file system size to 16TB. File system growth in an online process and will automatically expand DR file system as well, VMware could care less, NFS datastore just got bigger.
That's a good idea honestly. If the raid group itself is 16TB, I guess thats true that I don't really have to make the filesystem fill that whole space up front. I can grow it later.
I was just counting my drives and it does look like we can make it happen. It will take some careful planning. I need to migrate machines slowly anyway so the replication can keep up with it.
yes, the tricky part will be removing DAE's from prod NX4, just need to make sure that all dvols that reside on RG in that DAE are no longer in-use. (if that DAE/drives are dedicated to NX4 datamovers)