Triage Advice


cmannes

Recommended Posts

I have a semi-regular occurrence, where streaming a lower-res (<720p) video via binhex-plexpass, causes my entire unraid server (6.8.2, although it occured in prior versions as well) to lock up and become completely unresponsive.  Even a keyboard/mouse/monitor directly connected is unresponsive.  I have to power-cycle the server.  It's always come back fine, and everything works as expected.  Till the next hard-lock.

 

And while a 'fix' from the community would be nice, what I'd really like to find out, is how to triage the issue.  I'm actually a 'FullStack' developer, so I've got experience in Docker & Linux.  But for some reason the Unraid running from ram bit messes with my head, and I can't figure out where to start. :)

 

So if anyone can offer a little guidance, I'd like to learn more and dig into what's going on.

 

Thanks!

Link to comment

Okay, finally had it crash and kept the syslog. 

In the log I can see on 2/9 when I restarted the server around 20:40ish.  Then at 01:00 there's a long sequence of

 

Feb 10 01:00:06 Cube kernel: veth40a15f2: renamed from eth0
Feb 10 01:00:06 Cube kernel: docker0: port 1(veth5b405b7) entered disabled state
Feb 10 01:00:06 Cube kernel: docker0: port 1(veth5b405b7) entered disabled state
Feb 10 01:00:06 Cube kernel: device veth5b405b7 left promiscuous mode
Feb 10 01:00:06 Cube kernel: docker0: port 1(veth5b405b7) entered disabled state

I believe that's when my internet went out over night, so the various docker containers were complaining.  Around 1:11 the internet came back, and then there's a gap in the log between 01:11 and 13:19 when it was restarted after it locked up around noon.

 

Feb 10 01:11:26 Cube kernel: IPv6: ADDRCONF(NETDEV_CHANGE): vethb30dfd9: link becomes ready
Feb 10 01:11:26 Cube kernel: docker0: port 10(vethb30dfd9) entered blocking state
Feb 10 01:11:26 Cube kernel: docker0: port 10(vethb30dfd9) entered forwarding state
Feb 10 13:19:16 Cube kernel: microcode: microcode updated early to revision 0x2f, date = 2019-02-17
Feb 10 13:19:16 Cube kernel: Linux version 4.19.98-Unraid (root@Develop) (gcc version 9.2.0 (GCC)) #1 SMP Sun Jan 26 09:15:03 PST 2020
Feb 10 13:19:16 Cube kernel: Command line: BOOT_IMAGE=/bzimage initrd=/bzroot
Feb 10 13:19:16 Cube kernel: x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point registers'

The latest diagnostic is attached.  In the meantime I disabled hardware acceleration in Plex, and am going to try to crash it later once the latest parity check completes.

 

Thanks,

cube-diagnostics-20200210-1650.zip

Link to comment

So I installed Kiwi Syslog Service Manager on my PC.  it's running on 192.168.40.154.  In unraid I set "Local syslog server" "disabled" "remote syslog server" "192.168.40.154".  I hit apply, and nothing ever shows in Kiwi.
 

I checked the windows firewall, kiwi has full access tcp & udp.

 

I did a ping both ways.  And unraid & windows can see each other.

 

I tried

 

echo -n "test message" | nc -u -w1 192.168.40.154 514

 

And nothing.

 

I switched unraid & kiwi to TCP.  Tried

 

echo -n "test message" | nc -w1 192.168.40.154 514

 

And nothing.

 

So, what is the trick to getting unraid to save to a remote syslog?  I searched this forum, and most articles are really old.  I tried reddit, similar issues.  No youtube videos. 

 

 

 

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.