Sporadic disk failures

[Jon4got2]

Hello.

Over the last couple years I seem to have a reoccuring "disk failure" that can usually be remedied by changing the drive to a different bay in my JBOD chassis and making a new config or rebuilding the disk.  This happens about every 2 to 4 months, but the worst part is once or twice it has happenned that before I get a chance to rebuild the disk or in the middle of it, another disk will fail and I'll lose data.  Again, these disk will work fine once moved to another bay.  My fairly new 14TB parity disk is now showing read errors, and I'm fairly certain I can move it to another bay and it will fix this read error, but it's now a reoccurring event that I'd really like to get to the bottom of.  I assume it's a hardware issue, but before I go spending a ton of money on a new backplane or something for no reason, I'd greatly appreciate some guidance on how to troubleshoot a problem like this that is so infrequent.

Thanks in advance.  This community has been a life saver and a wonderful resource.

tower-diagnostics-20221025-1009.zip

[JorgeB]

Disk dropped offline and re-connected, this is usually a power/connection problem, you should also update the LSI firmware since that's free and there's a small chance it might help.

[Jon4got2]

Thanks @JorgeB.  I was pretty sure I had updated the firmware, but I went through the firmware upgrade process (linked) again just to be safe, and it was on the same version.  I have 20 hard drives and an 800 watt GameMax PSU, so I believe I have adequate power.  The chassis is an inexpensive chinese unit, so I'm wondering if the backplanes are not distributing the power adequately.  Or could be I need to solder these power connections?  Anything else that I can check/test?

[JorgeB]

You have firmware 16.00.10.00, current one is 16.00.12.00 IIRC, the one on the guide is to show how to update, not to show the latest firmware.

[Michael_P]

13 hours ago, Jon4got2 said:

distributing the power adequately.

 

Also, eliminate any splitters and don't hang all the drives off of one line back to the power supply

[Jon4got2]

6 hours ago, JorgeB said:

You have firmware 16.00.10.00, current one is 16.00.12.00 IIRC, the one on the guide is to show how to update, not to show the latest firmware.

I did go through those instructions which brought me to this page which only had the 16.00.10.00 firmware.  I'll look for that other firmware, but if you have a different link I'd appreciate it.

[JorgeB]

15 minutes ago, Jon4got2 said:

which only had the 16.00.10.00 firmware.

Yep, looks like the newer firmware was pulled, so keep that one for now.