VSAN Proof of Concept
-
I've seen the buzz about it and watched the webinars. I'm excited about VSAN 2.0 and would like to play with it (not just using HOLs). But I'm not in a position where I have 3 hosts (let alone 3 hosts on the VMWare and VSAN HCL) with a mix of spinning disks and SSDs to do any kind of proof of concept for it.
We use the Essentials bundle of VSphere 5.5 (3 hosts with all local storage and OBR 10) with Veeam for backups. For the most part, we can hit our acceptable downtime window of 1-2 hours. As we look to potentially go paperless out in our manufacturing shops in the next year or so (and thus shrink the acceptable downtime window to 30-60 minutes), I am thinking VSAN may be a good solution to give us an IOPs boost and some additional data resiliency. We might even look to go to Essentials Plus as well.
We're not dying for IOPs right now, but if everything was faster, users and execs become happier people. And if we had a little more storage resiliency, that makes it even better.
Again, we would only do this if business needs require it. But my question is more centered around how people get the gear to fully test out VSAN and do a proof of concept before they deploy it?
-
@NetworkNerd said:
Again, we would only do this if business needs require it. But my question is more centered around how people get the gear to fully test out VSAN and do a proof of concept before they deploy it?
Bigger companies that care about testing maintain labs with lots of hardware to test this stuff. That's the price paid for all kinds of necessary testing.
-
@scottalanmiller said:
@NetworkNerd said:
Again, we would only do this if business needs require it. But my question is more centered around how people get the gear to fully test out VSAN and do a proof of concept before they deploy it?
Bigger companies that care about testing maintain labs with lots of hardware to test this stuff. That's the price paid for all kinds of necessary testing.
And I am certain their internal IT is much larger than our 4-person team.
-
Cost out a basic lab that'll let you test this & present a plan to manglement. ROI will be iops boost, data resiliency and a place where you can learn, grow your skills and test things out. Look for used gear / servers. In the USA y'all have some SICK deals on ebay that make me (Canadian) cry like a little girl. If you want some warranty, hit up xbyte - they don't suck.
-
You might be able to get demo gear we have done this with high dollar SANs in the past.. Low end not so much. We had to compare it to equallogic just because management wanted to check out cheap. Well they brought in a demo. They didn't leave it with us and had a sales guy use it to demo for a few hrs. Noticed it was getting insane IOPS. Yep, figured out they use lots of small size stuff that fits with in the RAM cahce to make it look good in demos. So whatch out for scams like that if you get demos. EMC left a nice unit with us for a good while. Are you looking at vmware's own VSAN or solarwinds? How many hosts? If just a few you should be able to use slightly older hardware and test it.
-
Also check with a IT leasing company.
-
@NetworkNerd said:
@scottalanmiller said:
@NetworkNerd said:
Again, we would only do this if business needs require it. But my question is more centered around how people get the gear to fully test out VSAN and do a proof of concept before they deploy it?
Bigger companies that care about testing maintain labs with lots of hardware to test this stuff. That's the price paid for all kinds of necessary testing.
And I am certain their internal IT is much larger than our 4-person team.
With a small team you are stuck leveraging: reviews, logic, anecdotal evidence, vendor reputation and guesswork. Just the nature of being small. Leverage what the big boys do, applying logic as to when what they do matters only at their scale rather than at yours. The way that SMBs compete is by not doing all of the expensive lab work and trusting the work of others. The way that enterprises compete is by doing their own research better than someone else does it and getting a technical advantage.
One is about blazing new ground, the other is about running agile and low cost.
