Equallogic Storage Pool - Replication

Question

I am very new at EQL replication and am trying to wade through the white papers. I have several questions I was hoping to have answered. But first a little about my setup. I have a PS6110XV (firmware 6.05) and a single Storage Pool and Volume/LUN serving an ESX HA Cluster. The volume was created (not by me) at its near max of 4.8TB (thick volume) with very little for snapshots (5%) and local reserve. Current in use is 2.6TB with just over 1.8TB free. I am attempting to understand my options of setting up this unit as a primary group to replicate to a PS4110E and am having issues. I first brought the PS4110E up in its own group and am able to plug it into the local iSCSI network of the PS6110XV. All is good in regards to connectivity at this point. However, Im struggling with the question of whether I need to evacuate my volume entirely on the PS6110XV and start over with a new volume allocation because the primary group fails for lack of space which I believe is the lack of space I have to allocate to the local reserve. 1. Do I need a Local Reserve on the Primary Group equal or greater than the 'in use' portion of the single LUN on this unit? 1a. If yes to 1; My current volume on the PS6110XV is showing 2.6TB in use, however, I was successful to get my HA cluster down to 1.7TB. However, the volume on the EQL LUN still shows 2.6TB from writes I previously performed. I have attempted the vm tools to reclaim space but now understand that is only geared towards thin provisioned LUNs. Do I have an option to reclaim that space on the thick LUN, which should theoretically leave me with enough free space in the LUN to perform the initial replication?  1b. If yes to 1 and no to 1a: Can I add a second member (in my case a much OLDER PS5000) array to the Primary Group into another Storage Pool and utilize that for the Local Reserve of my PS6110XV LUN just to accommodate the replication? 1c. If yes to 1, and no to 1a and 1b: Do I need to evacuate my PS6110XV and start over and be smarter with my allocation and reserves to ensure I consider replication from the beginning?  2. Looking forward: Once I get this setup I will locate the secondary group over the WAN. I have plenty of bandwidth but am wanting to confirm a few things. I understand I can not NAT the traffic. Which tells me I have the single option of setting up a VPN or going with a L2 handoff. Is that correct? What I am mostly struggling with here is whether the default gateway of the management NIC or Group needs to be reachable to the remote location. I don't know if I am articulating myself well here and am fairly strong with the networking side but believe I have to setup proper routing (not utilizing NAT) within the iSCSI networks correct across a WAN link? Thank you for any advice/information here. I truly appreciate it. Jk

Jeremy.Kurth · Answer

I did exactly as Don said. I modified the volume settings to a thin provisioned volume. I then had to reduce the free space using ESX steps. Once that was complete I had enough free space on the volume to perform the replication to the 4110. It is currently in progress. One thing this tells me it to be cautious in sizing your PS SAN's if you will be doing replication. In short, it sounds like you should have at least 50% free space to utilize and configure replication.

Thanks again Don!

Jeremy.Kurth · Answer

Hi Don,

thank you so much for responding. Are you able to provide a clear and supported document to convert a thick volume to a thin one? Mostly what I find is related to converting a thin to a thick...

Again, I truly appreciate the assistance.

Jeremy

EqualLogic

Was this post helpful?