Proposed All SSD Build for VM SAN

[HoneyBadger]

Here are the results fromthe same host and same VM but targeting an NFS datastore. Significantly lower performance on the high queue/thread random tests.

View attachment 30592

If you haven't already, try a new dataset with recordsize=16K instead of default 128K over NFS.

 

[Ctiger256]

A few things here. I realized that I had way underprovisioned the vCPUs on my test VM which was bottlenecking the 8 thread test. Re-ran with 8 vCPUs and the numbers are much better. Here they are on iSCSI:



And on NFS (Now set at 16k recordsize):



The read performance now is about the same between iSCSI and NFS. But the write performance is a lot lower on NFS, particularly on the highly threaded writes. I thought maybe this was a sync issue. But both the iSCSI zVol and the NFS dataset have sync set to "Always" So i don't really know how to explain that.

On the iSCSI VM, I played with different thread and queue counts to see if I could push the speed higher but I seem to be maxed out on the 8T/8Q test--that comes out to roughly 100k IOPS. These drives - Samsung 883 DCT are spec'd at 98k IOPS read/28k IOPs write. I'm running 2 wide mirror/4 devs so I think that explain my writes maxing out at around 100k IOPs (~28k x 4). Not sure why I can't get higher reads. Maybe I'm not understanding ZFS Raid on random IO.

 

[Ctiger256]

[I'm going to move this discussion over to a new thread as it's moved away from "build" and onto "performance testing"]

 

[uberwebguru]

Wrong CPU. The 2637v4 is a quad core 3.5GHz-turbo-to-3.7 with 15MB cache that's designed for dual-socket operation and costs around $1000.

The 1620v4 is almost the same CPU except it turbos to 3.8 and only has 10MB cache, costs a third as much ($294). The 1650v4 is actually a sweet CPU, six cores 3.6GHz-turbo-to-4.0, around $600.

The one downside to the E5-16xx CPU's is that they do not support LRDIMM. This probably isn't a big deal.

Having built something like this but with HDD's, I'll also note that I had a hard time getting the CPU to do much more than yawn even under fairly heavy load. My *impression* is that the 1620 would be fine, but once you're paying for all the other components, a few hundred extra for more cores and the slight speed bump is really nothing.

Can one use E5-1600 v4 for CPU that support E5-2600 v4?
I only see support for E5-2600 v4 and not E5-1600 v4, so not sure

 

[jgreco]

Can one use E5-1600 v4 for CPU that support E5-2600 v4?
I only see support for E5-2600 v4 and not E5-1600 v4, so not sure

A Supermicro "S" series is typically a "S"ingle socket affair and so should be able to use either the 16xx or 26xx. The motherboard in question is an "S" series:

MBD-X10SRH-CLN4F

And per Supermicro

X10SRH-CLN4F | Motherboards | Products | Super Micro Computer, Inc.

X10S, X10SR, X10SRH, X10SRH-CLN4F

www.supermicro.com

1. Single socket R3 (LGA 2011) supports
   Intel® Xeon® processor E5-2600 v4†/ v3 and E5-1600 v4†/ v3 family

So I believe I'm correct.

 

[uberwebguru]

A Supermicro "S" series is typically a "S"ingle socket affair and so should be able to use either the 16xx or 26xx. The motherboard in question is an "S" series:



And per Supermicro

X10SRH-CLN4F | Motherboards | Products | Super Micro Computer, Inc.

X10S, X10SR, X10SRH, X10SRH-CLN4F

www.supermicro.com

1. Single socket R3 (LGA 2011) supports
   Intel® Xeon® processor E5-2600 v4†/ v3 and E5-1600 v4†/ v3 family

So I believe I'm correct.

Dell PowerEdge R730xd?

 

[jgreco]

Dell PowerEdge R730xd?

Check with your Dell sales rep.

 

[HoneyBadger]

Dell PowerEdge R730xd?

Nope. I think the tower servers can do the single or dual-socket CPUs, but the rackmounts are limited in the BIOS/EFI to E5-2600 series.

 

[uberwebguru]

Nope. I think the tower servers can do the single or dual-socket CPUs, but the rackmounts are limited in the BIOS/EFI to E5-2600 series.

yeah what i thought...i havent seen replacing E5-1600 v4 with E5-2600 v4 in same server model
just wanted to confirm am not missing out on anything