Hello! At this point, we face with a problem that ESXi iSCSI Initiator performance is bounded by CPU and limited by the performance of one core on ESXi host. In our lab, on 4K random reads we got only 150000 IOPS. Currently, we are looking for the possible solution of it. On the other hand, we have...

Ok I'll give it a shot with the NVMe cards and post results. Is the Linux distro soon to be beta? It would be interesting to compare.

-D

This has been discussed before, actually. There are apparently issues with the Windows stack that cap out performance. I guess that's why they're moving to linux?

We had to add Pernix on top of SW to make it perform like we thought it needed to.

-D

Any news yet? I'm not sure it really matters for us ATM, since Pernix is killing it on the frontside. It would be nice not to have to pay that renewal.

-Darren

Did you guys ever figure this out? I have a similar issue. Testing the SAN Array locally pushes 2GBps + on both read and write but connecting from an ESXi 6.0u2 host over 10Gbe I can only get about 250MB-300MB read and write. Not as of recently. Higher bandwidth and IOPS are both an issue. Apparent...

I've actually gotten some decent numbers shutting off all the caching off, and using a minimal amount of RAM for SW, and installing PERNIX on the front side. This really is a far cry from the original intent for this software. Hoping this gets fixed soon. The speed of drives have gone nuts in the la...

Here's a whole story: 1) Microsoft storage stack is a very old code, it was written and designed when underlying storage was slow. As a result Microsoft isn't going to deliver all possible IOPS from the single very fast NVMe card. There are a couple of workarounds: a) Disable I/O scheduler. There's...

That's a good link, thanks. I'm relying right now on the SW guys to nail this down. I feel as that may be squeezing the final IOPS out of the system. I'm only getting 20% of the native IOPS, so there must be something HUGE bottlenecking the system.

-D

May I ask how did you benchmark the physical storage? As far as I know, it is a bit complicated to simulate writes and reads directly on SSD drive. I think it’s just a matter of inbuilt cache which we cannot go around in doing so. Sure. We run IOmeter from the SW VM to get a base. Then we share out...

I've been working with the SW guys for the better part of 6 months trying to get some PCIE flash cards / SSD's running properly in a hyperconverged environment. I've tried many different combos, and even setup a physical SW server with some LSI Nytro drives. Basically, in a hyper converged environme...

There should be sort of log with flat storage (image files, LSFS are immune to this issue) to prevent us from not knowing what node has most recent data, that's referred as "amnesia issue". See: https://docs.oracle.com/cd/E19575-01/821-0518/concepts-42/index.html You can absolutely use fl...

I still don't understand how the L2 cache in WB is any more susceptible to data corruption than the WB ram cache on a regular store. If they are both mirrored on both nodes, where's the issue? Maybe one of the SW guys can explain that, since I'm still hoping I can use these FusionIO cards to help wi...

Did you ever fix this? We are running into the same exact situation.

It works as expected. OVF has a huge disclaimer. DO NOT USE TO EVALUATE PERFORMACE https://www.starwindsoftware.com/registration-starwind-virtual-san "For prompt deployment and functionality evaluation, StarWind Virtual SAN is available as a pre-configured VM. It allows you to install the soft...

OK we figured out part of the issue. The SW tech who did the install did not turn on Jumbo Frames inside the 2012 machines. We flipped them on, and now the system is useable, although some serious tweaking needs to be done. Speeds are still poor. The whole OVF install was really not smooth. I'd reco...