Potential HDD Failure? and Unraid slowly dies becoming unresponsive ...

[bombsngravy]

I've been running unraid for a bit now and haven't had any problems that weren't self-induced. Recently though Unraid will slowly start dying ... it usually starts with shares not being accessible outside of unraid and/or with specific dockers (plex for example), it will then become unreachable via IP and unresponsive at the machine.

 

I enabled syslog to get a more complete picture, it appears that one of my nvme's is failing ... but what follows doesn't make much sense to me, in the syslog attached the issues start occurring around 1843 ... at 2148 something happens ... this is when loss to shares/plex occurs and then it slowly dies after I access it at 0100... 

 

In the interim, I've disabled the nvme that it calls out in the log ... but don't know if i have other problems.

 

Thanks!

 

bskycc-diagnostics-20240530-0618.zip

[JorgeB]

Pool filesystem is going read only with

write time tree block corruption detected

This usually means a RAM issue, start by running memtest

[bombsngravy]

3 minutes ago, JorgeB said:

Pool filesystem is going read only with

write time tree block corruption detected

This usually means a RAM issue, start by running memtest

@JorgeB gotcha, I assume that's only at boot ? I'm currently remote from the system, so it'll be a bit before I can get back to it ... 

[JorgeB]

Yes, it an option on the boot menu

[bombsngravy]

9 hours ago, JorgeB said:

Yes, it an option on the boot menu

This is where it’s at right now - does it need to go to 100%?

[JorgeB]

Let it complete at least one pass, also note that memtest is only definite if it finds errors, but if it doesn't after a pass you can try recreating the fileystem, or try zfs, if issues persist there could still be an underlying hardware problem.