Hi All,

Over the last week I have seen my Unraid crash on average every 26 hours .

I ran a Memtest for 2/3 hours and no errors showed. I have gone through the logs many times and cannot see what is causing this. I have looked for trends in access, CPU usage, etc but cannot see any.

The WebGUI is not available, cannot SSH and console does not respond in a timely manner (I can type the login name but the login times out after 60seconds and never in time for me to log in). Additionally all dockers are unavailable and unresponsive.

Can anyone see anything I am missing?

Thanks!

GK

gk-unraid-diagnostics-20250817-1412.zip

I would suggest that you set up the syslog server and post up the syslog from after the next crash. Post it in a new post in this thread so that we know you have added additional information. (Do NOT edit your original post to add it!!!!)

Hi Frank,

Is this not available in the syslog-previous file in the diagnostics?

I had the server set up to mirror to flash prior to running diagnostics, but I 100% could be missing something or have gotten this wrong.

Thank you for the reply! :)

Hi Frank,

I've enabled the "Local syslog server" now as I didn't realise this was needed as well as the "mirror syslog to flash" option.

If the machine crashes again I will post new diagnostics.

Thanks,

GK

Edited August 17, 2025Aug 17 by gearoidocathain

3 hours ago, gearoidocathain said:

I didn't realise this was needed as well as theirror Syslog to Flash option.

Problem is that the diagnostics file is a snapshot of the system at the time when you call for it to be run. It can never be generated after the crash has occurred. Normally, Unraid stores the syslog in RAM with the rest of the OS files. Thus it is lost when you reboot the server. What the syslog server does a real-time copy every entry that is written to the syslog to 'hard' storage. Sometimes, it will point to what is causing the problem. Some times, it provides no clue but even this is a help in telling us where to look next for the cause.

I seem to recall that there is a switch somewhere that will write copy the entire syslog to flash as a part of the 'normal' shutdown process but, by definition, a crash is not a normal shutdown!

The syslog file generated by the syslog server is not included as a part of diagnostics. Probably, it does not always always exist and does not have a fixed file location. (BTW, you don't want to activate the mirror to flash option when you don't have a problem as it can result in 1000's of writes to your flash drive. And flash drives do have a limited number of write-cycles before failure!)

Edited August 17, 2025Aug 17 by Frank1940

Also make sure this has been taken care of:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/page/2/#findComment-819173

Hi Jorge,

Thanks for the link!

I had found that article previously and checked my setup to make sure there were no conflicts. The server has been running for approx 3 years with no issues previous to this and I have not changed or updated the BIOS since the build so I don't believe that this is causing my issues.

Thanks,
GK

Hi @Frank1940 ,

There is a "syslog-previous" file in the diagnostics zip. I think that this is the syslog right up to the point that the machine crashed as it has the timestamps up to around the time the machine failed? All of the rest of the files however are generated from the current machine state when diagnostics command is run.

I will keep trying to catch the machine as it runs into difficulty and run the diagnostics command then but its not easy :)

Thanks,

GK

Edited August 18, 2025Aug 18 by gearoidocathain

Of course it's now behaving itself and hasn't crashed today.... I'm keeping a close eye on it though.

15 minutes ago, gearoidocathain said:

Of course it's now behaving itself and hasn't crashed today.... I'm keeping a close eye on it though.

Keep in mind the environmental effects of temperature. If you can verify fans are operating correctly. Make sure that dust and dirt are not clogging vents and heat sinks.

Hi Frank,

This was my first thing to check as we have had a bit of warm weather recently but it has been nowhere near as warm as other times of the year. All vents etc are clear and fans are working.

Unfortunately (or fortunately?) the machine is running still today and there hasn't been a crash in over 2 days now. I'll update if anything happens but right now it's frustratingly stable.

Thanks,

GK