Has anyone tested the newest DELL PERC H730P firmware 25.5.4.0006 ?
This version was recently certified by VMware
VMware Compatibility Guide - I/O Device Search
I have concern how the drive cache change may impact vSAN storage performance...? Are they referring to HDD or SSD? I have the Intel SATA SSDs S3710 and S3610.
Fixes & Enhancements
Fixes
-Changes default value of drive cache for 6 Gbps SATA drive to disabled. This is to align with the industry for SATA drives. This may result in a performance degradation especially in non-Raid mode. You must perform an AC reboot to see existing configurations change.
I called Dell and talked to support, he asked the tier 3 engineer but answer was not helpful either - "may affect ssd drives", at last he told that is possible to enable / disable this option from the BIOS level...
Anyway, I've decided to go ahead and upgrade one of hosts in my cluster. I've installed the following packages:
Firmware:
- BIOS 2.7.1
- Backplane Extender 3.35
- iDRAC with Lifecycle Controller 2.52.52.52
- NICs 18.3.6
- PERCs 25.5.4.0006
- SSD DL2D
Drivers:
- PERC lsi_mr3-7.703.15.00-7534889
- NICs ixgbe-4.5.3-2494585-6925533
I put the host into production. The host is stable and working fine since last night.
Thank you for this feedback marnow!
Any warnings in the vSAN Health check in regards to the firmware version?
Bonjour,
I upgrade my cluster
Firmware:
- BIOS 2.7.1
- Backplane Extender 3.35
- iDRAC with Lifecycle Controller 2.50.50.50
- NICs 18.3.6
- PERCs 25.5.4.0006
- Full SSD
Drivers:
- PERC lsi_mr3-7.703.15.00-1OEM.650.0.0.4598673
I have an ESX that has messages in the "healt VSAN" warning and completely degrades the performance of the infra. I had to put it in maintenance mode for the moment.
If anyone has an idea I'm interested
Thank
Guys,
I have updated so far 3 hosts in my 8 nodes cluster. I'm not observing any negative performance impact. All HC is green.
roman79 No warnings, the 25.5.4.0006 FW version and driver is added to the HCL DB, Health Check reporting all green status.
Thank you marnow! Time to upgrade then
Hello marnow,
Just a note to clarify: you wouldn't really get an indication of performance degradation from the Health check unless it was severe enough to cause potential issues that there are tests for(e.g. disks going offline or congestion high enough to trigger the alert or warning for this test). Also, as this firmware version is on the vSAN-HCL and has been added to the JSON file that the Health check for hardware compatibility it's not going to trip that test either.
The appropriate method of achieving performance comparison would be to compare before and after (upgrading firmware on all nodes/making any other changes) outputs from vSAN Observer or by comparing the performance graph outputs in the Web Client while the cluster is under the same/similar workloads.
Bob
TheBobkin you are absolutely correct! roman79 asked for HC so I responded directly to him...I've checked the performance graphs before and after and with my typical workload I don't see a difference. I will perform another batch VMs migration with Storage vMotion next week and I will be closely watching the graphs, before the upgrade I was reaching about 35k IOPS on vSAN Backend.