[UNRAID 6.10 RC2] Server Shuts Down or Hard Locks After ~2 days

Not necessarily your problem, but all systems are far more stable when using the same memory sticks.  If you can't do that, then you really have to ensure that the CAS timing is identical between the two sets.

 

Either way, you want to run a memtest for at least a pass if only to rule out a bad stick (won't rule out the above though)

 

Now, if it actually shuts itself off then you're at either a power problem or an overheating problem.

Interestingly, this RAM has been in my system for 1-2 years. I did run a memtest last weekend and no errors were reported.

Im using Netdata and temps seem stable. Im wondering if its the power supply, but curious why psu would fail only after a couple days use. 

@Squid Is there a sort of stress test/cpu benchmark I can run in unraid to see if a specific component is dying out? Im thinking prime95 like tool... 

Running prime 95 via docker no issues....

Heres a pic of a crash

Enable the syslog server and post that log after a crash.

@JorgeB attached

tower-diagnostics-20211116-0924.zip

1 minute ago, aceofskies05 said:

attached

diagnostics only contains the current syslog since the last reboot, not the sylog from

58 minutes ago, JorgeB said:

the syslog server and post that log after a crash.

 

Why are you installing so many (all?) packages from DevTools and NerdPack? Do you actually use all of those? Do you even know what many of them are? I recommend not installing anything you don't use regularly.

 

 

I've had the syslog enabled for over  a year , that diagnostic log is after crash.... Is there something else you need? 

 

1 minute ago, aceofskies05 said:

that diagnostic log is after crash

 

40 minutes ago, trurl said:

diagnostics only contains the current syslog since the last reboot

so there is nothing there from before the crash. You need to get the syslog from wherever you told syslog server to save it and post that

Ah im following... thanks. Attached is the LOG before the crash..

Also nerd plugin has a check all button, must have click that on accident one day and enabled all plugins. I fixed that.

tower-diagnostics-20211115-1110.zip

5 hours ago, aceofskies05 said:

Ah im following... thanks. Attached is the LOG before the crash..

Maybe you understand, but you still attached a diagnostic instead of a syslog😉

Isnt the syslog in the Diagnostic log? Well, any who I've attached the syslog. Let me know if there's anything else you need. Appreciate the help

syslog.txt

22 minutes ago, aceofskies05 said:

Isnt the syslog in the Diagnostic log?

9 hours ago, trurl said:

diagnostics only contains the current syslog since the last reboot

23 minutes ago, aceofskies05 said:

I've attached the syslog

And that is also the current syslog since the last reboot. We want

9 hours ago, trurl said:

the syslog from wherever you told syslog server to save it

 

Got it. 7ths times a charm 😄. Here's the syslog from the mirror to flashdrive. 

syslog

Please test your system in safe-mode (no plugins installed).

Im getting random reboots now. I managed to catch it in the sys log. I logged in at 12:39:26 then it random rebooted right at 12:42:19, it then became available at 12:46:22
 

Dec  3 00:43:15 Tower sshd[27118]: pam_unix(sshd:session): session closed for user root

Dec  3 12:39:26 Tower webGUI: Successful login user root from 192.168.1.170

Dec  3 12:42:07 Tower kernel: veth4f47806: renamed from eth0

Dec  3 12:42:07 Tower kernel: br-4d8c83721e50: port 3(veth69d1485) entered disabled state

Dec  3 12:42:08 Tower kernel: br-4d8c83721e50: port 3(veth69d1485) entered disabled state

Dec  3 12:42:08 Tower kernel: device veth69d1485 left promiscuous mode

Dec  3 12:42:08 Tower kernel: br-4d8c83721e50: port 3(veth69d1485) entered disabled state

Dec  3 12:42:16 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered blocking state

Dec  3 12:42:16 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered disabled state

Dec  3 12:42:16 Tower kernel: device veth6996f31 entered promiscuous mode

Dec  3 12:42:16 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered blocking state

Dec  3 12:42:16 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered forwarding state

Dec  3 12:42:16 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered disabled state

Dec  3 12:42:19 Tower kernel: eth0: renamed from vethe963a49

Dec  3 12:42:19 Tower kernel: IPv6: ADDRCONF(NETDEV_CHANGE): veth6996f31: link becomes ready

Dec  3 12:42:19 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered blocking state

Dec  3 12:42:19 Tower kernel: br-4d8c83721e50: port 3(veth6996f31) entered forwarding state

Dec  3 12:46:22 Tower kernel: Linux version 5.14.15-Unraid (root@Develop) (gcc (GCC) 11.2.0, GNU ld version 2.37-slack15) #1 SMP Thu Oct 28 09:56:33 PDT 2021

Dec  3 12:46:22 Tower kernel: Command line: BOOT_IMAGE=/bzimage initrd=/bzroot pcie_acs_override=downstream,multifunction iommu=pt

Dec  3 12:46:22 Tower kernel: x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point registers'

Dec  3 12:46:22 Tower kernel: x86/fpu: Supporting XSAVE feature 0x002: 'SSE registers'

Dec  3 12:46:22 Tower kernel: x86/fpu: Supporting XSAVE feature 0x004: 'AVX registers'

Dec  3 12:46:22 Tower kernel: x86/fpu: xstate_offset[2]:  576, xstate_sizes[2]:  256

Dec  3 12:46:22 Tower kernel: x86/fpu: Enabled xstate features 0x7, context size is 832 bytes, using 'compacted' format.

Dec  3 12:46:22 Tower kernel: signal: max sigframe size: 1776

Dec  3 12:46:22 Tower kernel: BIOS-provided physical RAM map:

I've fresh installed unraid with 6.9-rc2. I'm narrowing down the error to Kernal "Dirty"

 

syslog

another pattern for crashes

Overnight stability reported by adding

"rcu_nocbs=0-15"