Disks gone, reboot, shares gone

[thany]

So I just noticed I couldn't reach my server via \\unraid. Brilliant. Went to the web interface, and saw to my surprize that all my disks had vanished. Poof. They worked fine yesterday, and nothing had happened in the mean time. Uptime is 7 days. They just vanished overnight. On HW failure, that may happen to one disk, but not all of them at the same time. Way too unlikely.

 

So I hit reboot. See if it re-detects them. Reboot does nothing - it just loads Boot.php, which is a totally empty page. So I went into the terminal, and entered `reboot` hoping that that will reboot the server. It did.

 

So the server is back up, and my array appears to be left undamaged. That's good. Sort of.

My shares are gone though. Sort of. They still exist in \\unraid, and dockers can see them, and I can see them in the terminal. The web interface just can't see them in the Shares tab. What gives? How can they exist and not exist at the same time? What Shrodingers magic is this?

 

Lastly, I need to know what transpired. I cannot reproduce the scenario, since I simply have no way of what went on. I need to know what went wrong. How can I find out?

 

I'm running Unraid 6.4.0.

 

TL;DR:
1) Please fix the reboot function in the Web UI.

2) How to find out what happened without knowing how to repro?

3) How to I get my shares back?

 

/edit

I just found that Add Share loads an empty page. So maybe fix that as well please?

 

/edit2

Also just found out 6.4.1 is released a while ago. Didn't get a notification of that. Maybe fix that as well please?

Anyway, the update doesn't fix any of the problems above. So I still need help.

[trurl]

If you have any adblockers, whitelist your server.

[bonienl]

Well, too bad you didn't create diagnostics before rebooting. I am inclined to say you experienced an out-of-memory situation. When that happens linux starts to kill random processes to free up memory, but this results in unpredictable behavior.

 

TL;DR:

1. There is nothing to fix

2 .Check your docker containers. A wrong path mapping results in writing to RAM memory = running out-of-memory

3. If it does happen again, post diagnostics before rebooting

[thany]

@trurl Huh. That's weird. Ad-blocker seems to disable those things. I wonder why. Or how?...

 

@bonienl I'll keep it in mind. I still believe unraid should *always* keep logs. After all, something might break or explode without warning, and then you have around or about 0 lines of log... I'm not sure why unraid opts out of keeping logs by default, but I can't imagine the benefits outweighing the total lack of being able to diagnose a situation. In short: I would've liked to know what happened *during* the failure, if any, not just *after*. I'm not sure how useful it's going to be, to have logs of a machine that is healthy, hardware wise.

[Frank1940]

17 minutes ago, thany said:

I still believe unraid should *always* keep logs.

 

IF you  install the 'Tips and Tweaks' plugin, it will save the logs on shutdown or reboot.  However, unRAID does not write logs 'on the fly'.  It stores them on its RAM disk.  If you don't want to install Tips and Tweaks, you can get the diagnostics files (which have a lot more troubleshooting information than the syslog) by going to 'Tools' page and then to 'Diagnostics' icon.  you can also get them by typing   diagnostics  on the command line if the GUI is not accessible.  

[trurl]

46 minutes ago, thany said:

Huh. That's weird. Ad-blocker seems to disable those things. I wonder why. Or how?...

Well known issue with adblockers. It triggers on something in the web page but it is a false positive. Surprised you haven't encountered the problem before. What version did you upgrade from?

[pwm]

21 hours ago, thany said:

I still believe unraid should *always* keep logs.

One issue with persistent logs is that if you store the logs on the flash you wear out the flash.

And RAM-stored logs doesn't survive a hang/reboot.

 

Maybe unRAID should make it easy to push logs to a log server on another machine. I have a number of flash-running devices that I have push logs to a centralized log server, which in itself is a quite tiny machine with a 2.5" drive. It records logs, MQTT data etc.

[Frank1940]

22 hours ago, thany said:

@bonienl I'll keep it in mind. I still believe unraid should *always* keep logs. After all, something might break or explode without warning, and then you have around or about 0 lines of log... I'm not sure why unraid opts out of keeping logs by default, but I can't imagine the benefits outweighing the total lack of being able to diagnose a situation. In short: I would've liked to know what happened *during* the failure, if any, not just *after*. I'm not sure how useful it's going to be, to have logs of a machine that is healthy, hardware wise.

 

1 hour ago, pwm said:

 

Maybe unRAID should make it easy to push logs to a log server on another machine. I have a number of flash-running devices that I have push logs to a centralized log server, which in itself is a quite tiny machine with a 2.5" drive. It records logs, MQTT data etc.

 

Perhaps one of you should start a thread in the "Feature Requests" sub-forum proposing that that a feature/option be added to unRAID which would allow the locations of the various logs files to assigned to some other device besides the RAM disk which might include a remote 'log server'.  

[John_M]

7 hours ago, pwm said:

Maybe unRAID should make it easy to push logs to a log server on another machine. I have a number of flash-running devices that I have push logs to a centralized log server, which in itself is a quite tiny machine with a 2.5" drive. It records logs, MQTT data etc.

 

I like the idea of that. I also like the idea of small boxes dedicated to particular tasks. I can imagine some people will want to run a virtualised log server though, hosted by their unRAID box...