ESXi 6.5.0 (HP build) on DL360e G8 crashing when Storage Adapter Cache installed.

[dehidding]

I have two HP DL360e Gen8 servers.

Both are configured identically:
2) Xeon E52430
80Gb RAM (40Gb per CPU)
B320i Smart Array controller (512mb Cache)

When the cache controller for the Smart Array is installed on one of the servers, VMware crashes.  The other server works fine with cache.  Trying another cache module - same result.  

So I'm trying to figure out what else I can try.

[e_espinel]

Hello.
The first thing would be to verify that there are no hardware errors in the internal logs of the equipment, especially regarding the disk controller
It would also be useful and recommended to update the firmware levels of the hardware.
Another detail is the driver you are using, attached link
https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX-5e3c1b48628348de87bd1a07ab

[dehidding]

iLO is not showing any errors regarding the disk controller.  I do get some iLO errors for using some DIMMs that are not HP Smart Memory.  If I remove those DIMMs and use only HP smart memory, the problem remains.

I'm using the most recent firmware that I could find for the hardware.  I'll check your link for the driver.

The HP custom ISO I'm using the the last one generated by HP for Gen8 servers...dated December 2019.  I would hope that it has the driver you have found.  How would I check?

[dehidding]

Refreshing my memory a little further.

The HP custom ISO I'm using the the last one generated by HP for Gen8 servers...dated December 2019.  However, both servers are running a slightly older version of the hpvsa Smart Array Driver (5.50-88OEM.550.0.0.1331820).  Using the version from the latest ISO slows disk access dramatically (10 times slower roughly).  This is a known issue with the hpvsa driver and has existed for at least 4 years.

Both servers are running the same ISO and drivers, but only one server is crashing when I install the cache module.

[e_espinel]

Hello.
In short :
According to what it indicates in both servers we have the same HP custom image and the same version of the driver for the Smart Array B320i
The Smart Array B320i comes with a 512MB cache by default.

How come you don't have the cache on one of these servers?

If you installed the HP iamge you must have the Smart Arrary CLI. The 5.50-88OEM.550.0.0.1331820 driver is from the ESXi 5.5, but HP recommended to use it for the ESXi 6.5 as well. So here are the instructions for the 2 versions.

From an SSH connection execute the following commands to get information about the smart array in each server and send the capture of what you see.

Show configuration
ESXi 5.5 -> /opt/hp/hpssacli/bin/hpssacli ctrl all show config
ESXi 6.5 -> /opt/smartstorageadmin/ssacli/bin/ssacli ctrl all show config

Controller status
ESXi 5.5 -> /opt/hp/hpssacli/bin/hpssacli ctrl all show status
ESXi 6.5 -> /opt/smartstorageadmin/ssacli/bin/ssacli ctrl all show status

Show detailed controller information for all controllers
ESXi 5.5 -> /opt/hp/hpssacli/bin/hpssacli ctrl all show detail
ESXi 6.5 -> /opt/smartstorageadmin/ssacli/bin/ssacli ctrl all show detail