• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

EMC VNXe Performance

T

toto3d

n00b
Joined
May 5, 2011
Messages
12
Just wondering if anybody had experience with the new VNXe 3100 yet, and how the performance was thus far. Will probably be purchasing one very soon and wanted a heads up on any issues/concerns with the device.

Thanks.
 
From a quick glance, the 3100 seems to be a real base model offering - no ability to run flash, iSCSI as it's only block protocol (no FC). It may be fine for some tiny implementations, but it certainly wouldn't do me any good anywhere in my environment.

ALL DEPENDS ON WHAT YOU USAGE/EXPANSION PLANS ARE - of which we have been given no requirements/usage plans.
 
My apologies...i listed my usage in another thread.

~2TB of data, 40 users, one Exchange, one SQL server for our accounting db, BES, and a couple file servers. nothing too intensive aside from those...and looking for growth. This would be our first virtualized environment.
 
Not sure how stringent your SQL and Exchange requirements are, but it may not be a very good substitute for local high performance SQL storage use. Keep in mind iSCSI has relatively high latency.
 
Also, how many spindles will be serving Ex/SQL? What kind? 15K SAS?
Will you be sticking to traditional LUNs (R5 data, R10 logs) for both Ex/SQL?
40 user data on seperate ATA disk? How many spindles? How many total VMs?
 
We are actually going to keep SQL a baremetal box with the db sitting on the SAN.

Would the NetApp FAS 2040a provide more performance? or are they pretty much similar in specs?
 
Right now we spec'd out all 15k SAS drives, about 14.4 TB raw data. all shares hosted on the SAN directly (CIFS), and about 11 -12 VMs.
 
toto3d said:
Right now we spec'd out all 15k SAS drives, about 14.4 TB raw data. all shares hosted on the SAN directly (CIFS), and about 11 -12 VMs.
Click to expand...

How many DAE's? SAS GB - 300's, 450's, 600's? Not meaning to throw tons of questions at you, but when you're dealing with low-end heads, there's not a lot of cache there and no flash to augment your spindles. During the peaks, you're going to be depending on how well you architected the disks benaeth your LUNs. If you do the still-preferred R10 for SQL/Ex logs, you might chew up a lot of spindles, just for logs to get your needed I/Os. Hopefully you did more, smaller drives vs. fewer, larger drives.

I've managed/deployed many NetApps including a FAS2050 - haven't gone any smaller than that (haven't messed with a 2040). I'd be concerned about expecting even a 2050 to do Ex/SQL and VMs. Only saving grace is Ex 2007/2010 don't demand anywhere near the I/O req's that 2003 did - hopefully you're not on 2003. That said, with enough mid-sized 15K SAS spindles in seperate aggregates from your VMs, you may be ok. Not sure how heavy your SQL gets hit. Is the Exchange only 40 or so mailboxes for your users?
 
Last edited:
2 DAE's, filled with 30 600GB 15K spindles...doing the math again that's about 18TB raw. 2 storage processors, and baremetal SQL with a ~15GB db. Running Exchange 2007 and yes, 40 or so mailboxes.

Thanks for the info - really helpful.
 
My mind says that the VNXe 3100 should be sufficient for our needs since we don't run any intensive applications except for Exchange. But my heart says I should just go big...with the 3300 instead which has more memory and optional 10Gb. Will I ever need that though?

Oh the choices....
 
toto3d: we currently have an NS20 with 2 DAE's. The first loaded with 300GB "high performance" drives and the other half loaded with 1TB "high capacity drives". Although we have fiber and iSCSI options, we have simplified our usage to VMware over NFS and CIFS usage. We have about 50 VM's served by a 5 host cluster. Each ESX host boots from local DAS (we originally had them booting from the NAS over fiber for a "baremetal" type approach but a rebuild could actually take longer due to fiber configuration).

We still do our 100 user Exchange on physical servers because we have some absurd users that think Exchange is a file storage solution and management is not concerned with educating them on their misunderstanding.

I mention all this to say, we are at the final stages of configuring a VNXe as a replacement and we are confident that the VNXe can handle our needs. We are in the 3rd year on our NS20 and we've found out that for roughly the same price as our annual support for our NS20, we can get an adequately configued VNXe...
 
Thanks. If you don't mind me asking, which model VNXe are you planning on getting? I'm currently leaning on going 3300 because of the better processor and the ability to switch to 10Gb down the line.
 
After a LOT of research and back and forth between EMC and DELL Equallogic, we've decided to buy a Scale Computing M4 cluster...

Interesting considering I'd never heard of Scale Computing 3 weeks ago. We were looking at, and were probably going to buy, an EMC VNXe 3300 with approx. 12TB space with everything except replication option. That was top shelf full of 300GB 15k and bottom shelf with a lot of SATA. EMC had made it pretty attractive in the $35,000 range ($5,000 recurring starting year 4). The Scale is 12TB usable but there are some very interesting features that make it worth looking at. For example, there's NO rebuild time (no hot spares either). The Scale option is going to run us $30,000 ($1,200 recurring starting year 4).

There are pros and cons with each but, for us, Scale won the day. If you're in the market, at least take a look.
 
You must log in or register to reply here.
Back
Top