L2 Cache not working on 1 Node?

Software-based VM-centric and flash-friendly VM storage + free version

Moderators: anton (staff), art (staff), Max (staff), Anatoly (staff)

Post Reply
midnight
Posts: 3
Joined: Mon Sep 09, 2019 2:17 pm

Mon Sep 09, 2019 2:42 pm

Hello,

Hopefully someone can help. I don’t think the L2 cache is working on the secondary node.

Brief info on setup (homelab) – potentially using this type of setup for our customers to replace hyper-v replication. Anyway………

I’m running VSAN Free on a 2-node hyper-v cluster. MPIO is set to least queue depth to help balance the iscsi between nodes. I’ve got VM’s running on both hosts – no problems.

I’ve got L1 caching set on both hosts to 48gb.

I’ve L2 caching set on both hosts to 230GB (256GB NVME drives).

The main storage image file is on the D: drive (both hosts)

The Flash-cache image file is on the E: drive (both hosts)

On the screenshots Node 2’s flash cache is showing 0% usage, all the others are fine (Node 1’s L1 & L2, & Node 2’s L1)

Is there a way to see if it is being used or if there are any logs or files I can check to see if there is a problem? The cache file is there on the E: drive on Node 2 and I can write to the folder without any problems.

Any ideas?

Many thanks
Attachments
Second Node
Second Node
Second Node.JPG (39.4 KiB) Viewed 2336 times
First Node
First Node
First Node.JPG (39.14 KiB) Viewed 2336 times
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Mon Sep 16, 2019 11:47 am

Hello midnight,
According to MPIO policy cache should be in use on both nodes.
Could you please collect the logs using this tool?
Please log a support case using the form.
midnight
Posts: 3
Joined: Mon Sep 09, 2019 2:17 pm

Mon Sep 16, 2019 9:44 pm

Hello,

Several days after submitting the original post I decided to change the MPIO policy to Round-Robin. The cache usage has now gone up to 4% - the others are still 100% though.

I'll keep an eye on it, but it looks like it might be working - albeit a little low compared to the others.

Thanks anyway.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Tue Sep 17, 2019 9:17 am

Hello,
Thank you for the update. If the issue persists, please proceed with the steps I have provided.
Post Reply