Daily crashing


Recommended Posts

All new hardware was used for this. I have a random crash anywhere from 24 hours to 72 hours.

 

I lose all network connectivity to the unit and have to force reboot it

 

Looking at the time stamps I think this is the kick off of when it starts.

 

Jan  3 06:02:09 Park-NAS1 kernel: rcu: INFO: rcu_sched self-detected stall on CPU

 

Hardware

Intel i3-10100

16gb ddr4 non ecc

intel x520-da2 nic.

6.9-rc2.

 

Edit: I added a diagnostics file while running. I'll try to capture one when it crashes but I lose all network connectivity when it does.

I'm looking at some of the settings in tips and tweaks and since it appears I lose network but the machine is still running and capture logs during that time, I'm going to turn off nic offload and see what that does.

 

syslog-127.0.0.1.log

park-nas1-diagnostics-20210103-2246.zip

Edited by donatom3
Addded diagnostics file
Link to comment
6 hours ago, JorgeB said:

Various call traces, including macvlan, those are usually the result of having dockers with a custom IP address:

I only had Plex using host. Had an issue when I first set it up a couple of months ago where using bridged wasn't determing local vs remote users well. Changed it to Bridge and it seems to work fine now (probably because I switched to linuxserver container).

 

Link to comment
14 minutes ago, donatom3 said:

I only had Plex using host. Had an issue when I first set it up a couple of months ago where using bridged wasn't determing local vs remote users well. Changed it to Bridge and it seems to work fine now (probably because I switched to linuxserver container).

Docker containers running in host or bridge mode are not known to cause macvlan call traces.

 

As noted, it is usually those containers that have their own IP address that cause macvlan call traces.  Have you assigned IP address to any containers on a custom network?

  • Like 1
Link to comment
30 minutes ago, Hoopster said:

Docker containers running in host or bridge mode are not known to cause macvlan call traces.

 

As noted, it is usually those containers that have their own IP address that cause macvlan call traces.  Have you assigned IP address to any containers on a custom network?

 
 
 

I used to have Nginx Proxy Manager running with it's own IP address on br0, it's not back to Bridged.. That was changed over 2 weeks ago. That should have been the only docker using it's own IP address.

Edited by donatom3
Link to comment

Just thought I would comment on using the forum. The reason I said

On 1/3/2021 at 12:46 PM, trurl said:

attach the complete Diagnostics ZIP file to your NEXT post

is because when you edit a previous post, the thread doesn't show anyone that there is new content. Since there were additional posts made to the thread after that, those showed up as new content, so I looked at the thread again. Since you didn't mention you had edited your previous post I thought perhaps you hadn't posted diagnostics even though you "liked" my post asking for them. Finally, I went back to your initial post and saw they were there. If there had been no additional posts to this thread it's possible the thread would have been ignored.

  • Like 1
Link to comment
12 hours ago, trurl said:

Just thought I would comment on using the forum. The reason I said

is because when you edit a previous post, the thread doesn't show anyone that there is new content. 

Thanks I realized I should have done that afterwards. I do have an update.

Crashed again. I could ping it but couldn't access ssh, web, or direct console.

 

I had the attachment on the screen. I decided to swap out the sata cable on dm-02.

Also noticed in the bios that the time was off and showing UTC time in bios.

 

I couldn't grab these diagnostics before the power down of course.

screenshot crash 1-5-21.png

park-nas1-diagnostics-20210105-1836.zip

Link to comment
  • 2 weeks later...
On 1/4/2021 at 9:10 AM, Hoopster said:

Docker containers running in host or bridge mode are not known to cause macvlan call traces.

 

As noted, it is usually those containers that have their own IP address that cause macvlan call traces.  Have you assigned IP address to any containers on a custom network?

 

What you said got me thinking. I did not have any dockers with a custom IP set but macvlan was still loaded like you said. So I just went into the docker settings and saw I still had custom br0 enabled. I disabled that and lsmod no longer shows macvlan loaded. I'll try for a few days like this.

 

Would the macvlan module being loaded be enough to cause the issue even if no docker was using a custom IP?

Link to comment

Just an update. I'm sitting on over 4 days uptime, keeping my fingers crossed but I feel good about this. Thanks Hoopster for pointing out MacVlan. I thought it wasn't in effect since I wasn't using a custom ip or br0 on any docker, but just having the br0 option checked in docker was enough to add it.

  • Like 1
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.