Yesterday was Day 3 of EMC World, and there were more great sessions packed full of technical information. Yesterday was also the last day of the exhibit hall being open so it was Apple iPad giveaway day as well, sadly I didn’t win one.
The first session that I want to recap here was the SAN meets NAS sessions that I attended. One of the big takeaways from EMC World was the technology that EMC is putting into all of the mid-tier products. This includes the EMC Celerra which is EMC’s Network Attached Storage product. The Celerra is basically as EMC CLARiiON with no fiber ports and a NAS connector on it, with a lite version of NaviSphere Manager running on it (unless you get the gateway only, and you then present LUNs from another fiber channel storage platform). What the FAST package lets you do is to have the hardware automatically move less used data from expensive storage to cheaper slower storage. This allows you to keep the data online so that your users can access it, but the access times will be just a little bit slower. Instead of having a 5ms response time it may have a 50ms response time, but just for the files which are older and haven’t been touched in a while.
When you setup FAST on your Celerra it defaults to 80% of the least active data will be moved to slower storage, but this number can be easily adjusted in the settings for the Celerra. As part of the Celerra FAST package you can even move the least active data to another device such as a Centera or Atmos device.
The second session was a session on getting Exchange 2010 setup and running on VMware’s vSphere 4.0.
VMware has shown both in their lab and their own production environment that they can run about 8000 users per virtual exchange server when you assign 8 vCPUs and 48 Gigs of RAM to each virtual server. Now for smaller virtual exchange servers VMware has these configuration guidelines.
|Number of Users||vCPUs||RAM|
Now do keep in mind that these numbers assume a 350 Meg mailbox size, so if you allow for a larger mailbox size than this you’ll need to adjust your numbers accordingly. Something to note that when putting the 8000 user Exchange server under load using Load Runner, only 2 vCPUs were running at 100%. While the recommended spec for an 8000 virtual user Exchange Server is an 8 vCPU virtual machine, a 4 vCPU virtual machine should work just fine.
One of the reasons that Exchange 2010 works so well in a virtual environment is because Microsoft has made some major changes in Exchange 2010 so that it requires a lot less disk IO that Exchange 2007. You’ll also notice that with Exchange 2010 (as well as older versions of Exchange and anything else really that has a well written database engine) that the more RAM you add to your virtual Exchange Server the less disk IO you’ll see as the Exchange Server is able to cache more and more data into RAM.
I spent a lot of the rest of the day milling around the exhibit hall looking at the products like the Dell and Cisco blade servers (which I’m looking at for our next VMware Server Expansion) as well as the new Dell 810 servers which can give you a quad socket server in a 2 U configuration. Both of which would give a much more dense solution than I’ve got now.
I also looked at the Xsigo virtual network/HBA configuration. It looks to be an kind of expensive solution to get into, but it would dramatically reduce the number of cables that I have to run to each server. Currently when I deploy a new VMware server I have to run 2 fiber cables and 13 Ethernet cables. This Xsigo solution would cut that down to two infiniband cables, each of which can carry 20 Gigs of bandwidth. This would reduce not only the cabling but the number of network switches that I have to buy (granted I’d be buying Xsigo modules instead). It’s an interesting solution, however Emulex has a couple of blog posts (“Candygram for Mongo” and “Where They At“) where they talk about the solution which make for some interesting reading.
Sadly in my walking around the show floor I managed to not win anything beyond the free t-shirts, pens, etc. Hopefully next time.
Last night was the traditional Wednesday Night party called the Beantown Bash. I’ve uploaded some pictures from the party (and some other stuff) to Flickr.
If you missed my other recaps…
There will be one more small wrap up tomorrow then that’s it until EMC World 2011.