As of last night at midnight, all was fine. I got my array status notification. But sometime after that my server crashed. I can't get to webGUI or SSH.

The errors on the monitor connected to the machine keep coming up every few seconds or so. See images below that my wife sent me (I am away at the moment).

 

I asked her to try to log in so she could shut it down via command line, but the errors popping up keep interrupting her login.

 

So, first question....is this a memory issue? Seems from some short googling that may be the case. I've ordered more RAM just in case.

Second, should I just pull the plug and abandon a graceful shut down?

 

[EDIT] - I just noticed this error in one of her other photos right after she apparently successfully logged in:

-bash: /usr/share/bash-completion/bash_completion: Input/output error

 

Thanks!

 

Edited September 18, 20232 yr by Mattaton

Did you ever discover the answer to this. Mine started doing the exact same thing this morning.

10 minutes ago, skulker said:

Did you ever discover the answer to this. Mine started doing the exact same thing this morning.

My issue ended up being that I was using macvlan for custom docker networks instead of ipvlan. See this thread:

Hopefully that helps you!

I'm using ipvlan and I'm not using dockers. Looks like I have some trouble-shooting to do. Thanks for the quick rpely!

1 minute ago, skulker said:

I'm using ipvlan and I'm not using dockers. Looks like I have some trouble-shooting to do. Thanks for the quick rpely!

Ah man. Sorry.

Maybe my new hardware/build fixed this issue, but I still had the issues in the other thread which ipvlan ultimately fixed. Thought the two problems were related, but maybe not.

Good luck with it!

I pulled the power plug and booted back up and its been running normally again the last couple of days (knock on wood). Whatever it was a power eset seems to fix it (at least for now). Thanks for the help!

Bumping since mine has started doing this twice this week.  Its been rock solid for along time..  and when this happens if i just hit reset it does not boot my USB, i have to pull it out and replug it before it will boot.. Im wondering if my USB is starting to fail.

25 minutes ago, RonneBlaze said:

Bumping since mine has started doing this twice this week.  Its been rock solid for along time..  and when this happens if i just hit reset it does not boot my USB, i have to pull it out and replug it before it will boot.. Im wondering if my USB is starting to fail.

Mine has done it again 3 times in the past month. And even though last time it was supposedly my use of macvlan on docker, this time I am set to ipvlan. So, it's not that.

 

What you describe with the USB is EXACTLY what happens to me. I have two USB ports on the front of the case and I just bounce between them. Very annoying. Hopefully we can get some help. But yeah I do wonder if it's the USB failing.

Did you (or anyone) find a root cause for this, and potential solution? I just started experiencing this, and it’s happened 3 times within the last week. Do I need new memory or replace boot USB as others alluded to?

Have you done memtest? 

Still happens periodically to me on entirely new hardware (other than the USB). It'll go like two months fine, and then pull this a couple times in one week. So I can't blame the memory with all new hardware. That points more to a USB issue, I suppose.

@Hella_captchas to answer your question, no. I haven't found a solution. Guess I'll try swapping the USB when I get a chance 

23 hours ago, trurl said:

Have you done memtest? 

Ran memtest today, 4 passes with 0 errors.  

 

I am also running a low profile usb stick.  While it hasn't died, maybe there is something to the heat it's absorbing?   I added a USB extender, and using a port from the rear this time.  

Edited April 25, 20251 yr by Hella_captchas
typo

Any updates on this error? I've been dealing with it for the past few days as well; server will run for about 24 hours fine, then no WebGUI, CLI/SSH, or access to Docker containers till a hard reboot and I remove/reinsert the USB boot drive. I've tried setting up syslog to write to the array and mirror to flash, but nothing meaningful writes since, once Exit Status 135 occurs, the array and flash drive are no longer accessible (checking through terminal on the actual device). I don't have another server setup, so writing externally isn't really an option.

 

For me, this seems to have started when I attached 2 external drives through USB and set them up as a new pool (my array is nearly full, I don't have room for more internal drives, and don't want to include USB drives in the main array for obvious reasons). I want to expand my array with larger drives (already purchased, just need testing), but I'm afraid this will happen during the resilvering process and that'll be yet another issue.

• ran memtest overnight (~9.5 hours); 12 passes; 0 errors; I'm fairly certain this is not a memory issue here)
• uninstalled & reinstalled plugins
• docker is set to use ipvlan

So, I guess the main question is: what does this exit status actually mean? Once we have that, we should have a better idea of how to fix it.

22 minutes ago, SlinkyBiest said:

I've tried setting up syslog to write to the array and mirror to flash, but nothing meaningful writes

zip the saved syslog and post it anyway, along with your diagnostics.

 

On 5/4/2025 at 10:24 PM, trurl said:

zip the saved syslog and post it anyway, along with your diagnostics.

 

Sorry for the delay, I was waiting for it to fail again. I started my machine in safe mode (plugins disabled) and it ran fine for a few days. Then, restarted it normally and uninstalled all plugins and it's been running with no issues for about a week; even after reinstalling a Community Applications and User Scripts since those are "crucial" to my current setup. I may start slowly reinstalling plugins once the new HDDs have been installed, but so far, it seems one of the plugins I had installed was the culprit.

 

It would still be nice to have a better error code (or a list of possible meanings somewhere) so users can more easily diagnose their systems on their own. It almost seems like 135 is just a catch-all - not knowing whether it's at least a hardware or software issue without digging heavily into logs is fairly problematic.