Is my RAID card or hard drive dying?

[veryoldgd]

I run FreeNAS as a VM with ESXi. It's got ECC RAM and HDDs connected via an IBM M1015 (flashed to LSI 9211-8i, and passthrough'd to the VM). It had been running pretty well until recently.

Basically, upon booting, I see a lot of:

Code:

(probe4:mpslsi0:0:4:0): INQUIRY. CDB: 12 00 00 00 24 00 length 36 SMID 1 terminated ioc 804b scsi 0 state c xfer 0
(probe4:mpslsi0:0:4:0): INQUIRY. CDB: 12 00 00 00 24 00 length 36 SMID 2 terminated ioc 804b scsi 0 state c xfer 0
(probe4:mpslsi0:0:4:0): INQUIRY. CDB: 12 00 00 00 24 00 length 36 SMID 3 terminated ioc 804b scsi 0 state c xfer 0

Rebooting the VM or the ESXi box does not work, I will have to shutdown the box and start it. Then it will work for a few minutes then stop working again.

I have a feeling that the RAID card is failing but I'm not entirely sure.

 

[Chris Moore]

I have a feeling that the RAID card is failing but I'm not entirely sure.

What you have is a SAS (Serial Attached SCSI) HBA (Host Bus Adapter) because flashing the card with IT mode firmware takes all RAID functionality out of the card. FreeNAS and ZFS are providing that function through software.

Then it will work for a few minutes then stop working again.

It does sound like a card failure. I had one fail on me and it was a simple matter of replacing the card and everything worked again. Since it is working for a bit and then stops, it sounds like it could be overheating. They do generate a good bit of heat. Do you have good air flow over the card heat-sync?

 

[veryoldgd]

I do not have a fan blowing directly towards the card. It is summer down here but at least recently it has been relatively cool. The card has served me for about 3 years and it has seen some hotter days but I really just started having issues with it from about...a couple of months ago probably.

Are the cheapo ones from Chinese sellers on ebay any good? I am going to China soon so might as well pick a few up. Local suppliers probably sell genuine brand new ones but it's bloody expensive.

Also I've noticed the disks will periodically spin up and make some weird noise. A speaker somewhere (or maybe mechanical noise?) would make a subtle low frequency low volume beep.

View Disk shows nothing :( and now even a cold boot doesn't seem to bring it online any more.

 

[Chris Moore]

I do not have a fan blowing directly towards the card. It is summer down here but at least recently it has been relatively cool. The card has served me for about 3 years and it has seen some hotter days but I really just started having issues with it from about...a couple of months ago probably.

Are the cheapo ones from Chinese sellers on ebay any good? I am going to China soon so might as well pick a few up. Local suppliers probably sell genuine brand new ones but it's bloody expensive.

Also I've noticed the disks will periodically spin up and make some weird noise. A speaker somewhere (or maybe mechanical noise?) would make a subtle low frequency low volume beep.

View Disk shows nothing :( and now even a cold boot doesn't seem to bring it online any more.

That's similar to what happened to me, on bare metal, and I just happened to have a spare card. If you can pickup another one, it would likely get you back in action

Sent from my SAMSUNG-SGH-I537 using Tapatalk

 

[brianm]

Check the firmware version of the controller card.
In my case flashing the card with LSI 9211-8i installed firmware 20.00.00.00 which produced a lot of errors' Flashing with LSI 9210-8i installed 20.00.07.00 which solved the problem.
You need driver version 21.01.00.00 and firmware version 20.00.07.00.

 

[veryoldgd]

Check the firmware version of the controller card.
In my case flashing the card with LSI 9211-8i installed firmware 20.00.00.00 which produced a lot of errors' Flashing with LSI 9210-8i installed 20.00.07.00 which solved the problem.
You need driver version 21.01.00.00 and firmware version 20.00.07.00.

IIRC I have 20.00.00.00 on the card. But I haven't upgraded FreeNAS since December 2015 so I don't think this is the issue.

 

[Ericloewe]

20.00.00 is very buggy, definitely update to .07.

 

[veryoldgd]

Okay this may sound weird. I think it's actually my hard drive bay or the power supply.

One day my SSD for the OS went missing in ESXi which had never happened before, and it was installed in a hard drive bay. I removed it from the bay and connected directly into the motherboard, and it came back.

This sparked me to remove my hard drives from the bay and connect directly to the HBA, boom they are showing up again.

Bad news is for some reason the EXSi is not passing through the HBA...need to fix that then put the system back into action for a while to be sure about my card.

FWIW the bay was powered by 3 Molex connectors from one Molex connector from the power supply using a splitter. It could be a bad practice. That's where my suspicion for the power supply (not necessarily the PSU) comes from.