I have 2 servers. "Main Server" one for VMs with GPU passthrough and a few low resources dockers. The other is the "Media Server" with lots of dockers, specially those that use GPUs for transcoding or AI. I use unassigned devices to mount some shares between them.

I woke up this morning and the Media Server was not serving the Samba shares. I go check and WebUI is not there at all. Not even Nginx serving error pages. SSH said connection refused or reseted by peer, I can't remember exactly. But the WebUIs of the docker containers appear to be fine. At least the 3 random ones I checked (Plex, Sonarr & Radarr).

I go to the WebUI of the Main Server and click on the share from Media Server that is still mounted according to the WebUI, hoping to see some error or any info. That causes the WebUI to stop responding. Opening a new tab and navigating to the WebUI now causes the browser to forever load a page that never arrives.

I SSH into Main Server and look around. Lot's of [Main Server hostname] key.dns_resolver: [Media Server hostname].LOCAL: No address associated with name error logs. lsof doesn't work cause of lsof: no pwd entry for UID 100 and lsof: no pwd entry for UID 508 and killing the processes started by those UIDs is not enough cause they get restarted immediately. Get tired of trying random things so I run diagnostics.

Shutdown both servers. Media Server via holding power button because I don't know how else. Power on both and generate diagnostics for Media Server. Here are both attached.

Can somebody help me shed some light on what has happened? So far I only noticed the Docker filesystem was at 90% so I increased it. But I doubt that made everything stop working BUT the dockers, right?

unblue-diagnostics-20250806-1147.zip unblue2-diagnostics-20250806-1222.zip

Edited August 19, 2025Aug 19 by BlueSialia

WebUI and SSH stopped working in Server 2 again. This time I didn't do any stuff in Server 1 so it doesn't freeze and I went directly to the syslog because each server sends the syslog to each other.

And the syslog of Server 2 has nothing relevant. Latest entries are from almost 12 hours ago and are totally normal:

Aug 18 10:52:47 UnBlue2 webgui: fileflows: Could not download icon https://raw.githubusercontent.com/revenz/FileFlows/master/icon.png?raw=true
Aug 18 10:53:02 UnBlue2 webgui: grocy: Could not download icon https://raw.githubusercontent.com/linuxserver/docker-templates/master/linuxserver.io/img/grocy-logo.png
Aug 18 10:53:03 UnBlue2 webgui: stirlingpdf: Could not download icon https://raw.githubusercontent.com/Frooodle/Stirling-PDF/main/src/main/resources/static/favicon.png
Aug 18 11:26:32 UnBlue2 monitor_nchan: Stop running nchan processes
Aug 18 11:28:31 UnBlue2 webgui: fileflows: Could not download icon https://raw.githubusercontent.com/revenz/FileFlows/master/icon.png?raw=true
Aug 18 11:28:46 UnBlue2 webgui: stirlingpdf: Could not download icon https://raw.githubusercontent.com/Frooodle/Stirling-PDF/main/src/main/resources/static/favicon.png
Aug 18 12:27:13 UnBlue2 webgui: Successful login user root from 192.168.10.21
Aug 18 12:27:42 UnBlue2 monitor_nchan: Stop running nchan processes

The latest warning is from over a day ago:

Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application frigate has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application grocy has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application paperless has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application qbittorrent has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application Rallly has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application romm has an update available for it
Aug 18 04:41:23 UnBlue2 root: Fix Common Problems: Warning: Docker Application swag has an update available for it
Aug 18 04:43:14 UnBlue2 root: Fix Common Problems: Warning: The plugin folder.view.plg is not known to Community Applications and is possibly incompatible with your server ** Ignored

The latest error is from almost a week ago:

Aug 13 12:51:45 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 0, async page read
Aug 13 12:51:45 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 0, async page read
Aug 13 12:51:45 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 0, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 0, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 0, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 4, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 8, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 16, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 32, async page read
Aug 13 12:51:46 UnBlue2 kernel: Buffer I/O error on dev md2p1, logical block 64, async page read

Edited August 19, 2025Aug 19 by BlueSialia

If there's nothing relevant logged, this can also be a hardware issue, one thing you can try is to boot the server in safe mode with all docker containers/VMs disabled, let it run as a basic NAS for a few days, if it still crashes it's likely a hardware problem, if it doesn't start turning on the other services one by one, including the individual docker containers.

It took almost 2 weeks for the issue to happen again. So I don't know how I'll be able to tell that it's not happening.

I could stop one of the 30 dockers running on it. Spend weeks without having a problem and I'll still doubt if it's fixed or if I was just lucky.

It's something I can try, and I appreciate the advice. But it's not very practical.

This issue keeps happening to me. No clear pattern. And this time no docker appears to function.

Sometimes the server reaches an uptime of almost a month, sometimes it lasts less than a week.

I plugged a HDMI display to check if something was displayed. It shows a GUI login screen but I cannot interact with it with a plugged keyboard.

Also, the HDD LED light of the case is flashing at regular intervals of 500 ms.

Any idea on what I can do to diagnose this? I have like 40 dockers in this server. Running in safe mode with all dockers disabled and start them one by one each month will take years. And most of those dockers I use daily.

These are the last 2 hours of logs as received on my other server:

Jan 12 06:00:01 UnBlue2 Plugin Auto Update: Checking for available plugin updates
Jan 12 06:00:13 UnBlue2 Plugin Auto Update: Auto Updating parity.check.tuning.plg
Jan 12 06:00:13 UnBlue2 plugin-manager: executing inline script: /bin/bash '/tmp/inline1-parity.check.tuning.sh'
Jan 12 06:00:13 UnBlue2 plugin-manager: creating: /boot/config/plugins/parity.check.tuning/parity.check.tuning-2026.01.09.txz - downloading from URL https://raw.githubusercontent.com/itimpi/parity.check.tuning/master/archives/parity.check.tuning-2026.01.09.txz
Jan 12 06:00:13 UnBlue2 plugin-manager: running: upgradepkg --install-new /boot/config/plugins/parity.check.tuning/parity.check.tuning-2026.01.09.txz
Jan 12 06:00:14 UnBlue2 plugin-manager: executing inline script: /bin/bash '/tmp/inline3-parity.check.tuning.sh'
Jan 12 06:00:22 UnBlue2 Parity Check Tuning: Versions: Unraid 7.2.0, Plugin Version: <unknown
Jan 12 06:00:22 UnBlue2 Parity Check Tuning: Configuration
Jan 12 06:00:22 UnBlue2 Parity Check Tuning: Array#012(#012    [parityTuningScheduled] => 1#012    [parityTuningManual] => 0#012    [parityTuningAutomatic] => 0#012    [parityTuningFrequency] => 0#012    [parityTuningResumeCustom] => 5,15,25,35,45,55 * * * *#012    [parityTuningResumeDay] => 0#012    [parityTuningResumeHour] => 0#012    [parityTuningResumeMinute] => 15#012    [parityTuningPauseCustom] => 0,10,20,30,40,50 * * * *#012    [parityTuningPauseDay] => 0#012    [parityTuningPauseHour] => 7#012    [parityTuningPauseMinute] => 30#012    [parityTuningNotify] => 1#012    [parityTuningRecon] => 0#012    [parityTuningClear] => 0#012    [parityTuningRestart] => 1#012    [parityTuningShutdown] => 0#012    [parityTuningMover] => 1#012    [parityTuningBackup] => 1#012    [parityTuningBackground] => 1#012    [parityTuningDocker] => 0#012    [parityTuningContainers] => #012    [parityTuningHeat] => 0#012    [parityTuningHeatHigh] => 3#012    [parityTuningHeatLow] => 8#012    [parityTuningHeatNotify] => 1#012    [parityTuningHeatShutdown] => 0#012    [parityTuningHeatCritical] => 2#012    [parityTuningHeatTooLong] => 30#012    [parityTuningLogging]
Jan 12 06:00:23 UnBlue2 plugin-manager: parity.check.tuning.plg updated
Jan 12 06:00:23 UnBlue2 Plugin Auto Update: Checking for language updates
Jan 12 06:00:23 UnBlue2 Plugin Auto Update: Community Applications Plugin Auto Update finished
Jan 12 06:07:34 UnBlue2 flash_backup: adding task: /usr/local/emhttp/plugins/dynamix.my.servers/scripts/UpdateFlashBackup update
Jan 12 06:30:03 UnBlue2 Docker Auto Update: Community Applications Docker Autoupdate running
Jan 12 06:30:03 UnBlue2 Docker Auto Update: Checking for available updates
Jan 12 06:30:10 UnBlue2 webgui: fileflows: Could not download icon https://raw.githubusercontent.com/revenz/FileFlows/master/icon.png?raw=true
Jan 12 06:30:27 UnBlue2 webgui: romm: Could not download icon https://raw.githubusercontent.com/rommapp/romm/release/.github/resources/isotipo.png
Jan 12 06:30:30 UnBlue2 webgui: stirlingpdf: Could not download icon https://raw.githubusercontent.com/Frooodle/Stirling-PDF/main/src/main/resources/static/favicon.png
Jan 12 06:30:38 UnBlue2 webgui: fileflows: Could not download icon https://raw.githubusercontent.com/revenz/FileFlows/master/icon.png?raw=true
Jan 12 06:30:39 UnBlue2 webgui: romm: Could not download icon https://raw.githubusercontent.com/rommapp/romm/release/.github/resources/isotipo.png
Jan 12 06:30:39 UnBlue2 webgui: stirlingpdf: Could not download icon https://raw.githubusercontent.com/Frooodle/Stirling-PDF/main/src/main/resources/static/favicon.png
Jan 12 06:30:39 UnBlue2 Docker Auto Update: Stopping lubelogger
Jan 12 06:30:42 UnBlue2 kernel: br-f687e37b0ff7: port 5(veth8bad4ae) entered disabled state
Jan 12 06:30:42 UnBlue2 kernel: veth07a687e: renamed from eth0
Jan 12 06:30:42 UnBlue2 kernel: br-f687e37b0ff7: port 5(veth8bad4ae) entered disabled state
Jan 12 06:30:42 UnBlue2 kernel: veth8bad4ae (unregistering): left allmulticast mode
Jan 12 06:30:42 UnBlue2 kernel: veth8bad4ae (unregistering): left promiscuous mode
Jan 12 06:30:42 UnBlue2 kernel: br-f687e37b0ff7: port 5(veth8bad4ae) entered disabled state
Jan 12 06:30:42 UnBlue2 Docker Auto Update: Stopping qbittorrent
Jan 12 06:30:47 UnBlue2 kernel: br-f687e37b0ff7: port 27(vethf53607d) entered disabled state
Jan 12 06:30:47 UnBlue2 kernel: vethe133876: renamed from eth0
Jan 12 06:30:47 UnBlue2 kernel: br-f687e37b0ff7: port 27(vethf53607d) entered disabled state
Jan 12 06:30:47 UnBlue2 kernel: vethf53607d (unregistering): left allmulticast mode
Jan 12 06:30:47 UnBlue2 kernel: vethf53607d (unregistering): left promiscuous mode
Jan 12 06:30:47 UnBlue2 kernel: br-f687e37b0ff7: port 27(vethf53607d) entered disabled state
Jan 12 06:30:47 UnBlue2 Docker Auto Update: Stopping readarr
Jan 12 06:30:51 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth60d325f) entered disabled state
Jan 12 06:30:51 UnBlue2 kernel: veth567e976: renamed from eth0
Jan 12 06:30:51 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth60d325f) entered disabled state
Jan 12 06:30:51 UnBlue2 kernel: veth60d325f (unregistering): left allmulticast mode
Jan 12 06:30:51 UnBlue2 kernel: veth60d325f (unregistering): left promiscuous mode
Jan 12 06:30:51 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth60d325f) entered disabled state
Jan 12 06:30:51 UnBlue2 Docker Auto Update: Stopping speakarr
Jan 12 06:30:54 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth0452dff) entered disabled state
Jan 12 06:30:54 UnBlue2 kernel: vetha3313a0: renamed from eth0
Jan 12 06:30:54 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth0452dff) entered disabled state
Jan 12 06:30:54 UnBlue2 kernel: veth0452dff (unregistering): left allmulticast mode
Jan 12 06:30:54 UnBlue2 kernel: veth0452dff (unregistering): left promiscuous mode
Jan 12 06:30:54 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth0452dff) entered disabled state
Jan 12 06:30:54 UnBlue2 Docker Auto Update: Installing Updates for lubelogger qbittorrent readarr speakarr
Jan 12 06:32:14 UnBlue2 Docker Auto Update: Restarting lubelogger
Jan 12 06:32:14 UnBlue2 kernel: br-f687e37b0ff7: port 5(vethe592bd0) entered blocking state
Jan 12 06:32:14 UnBlue2 kernel: br-f687e37b0ff7: port 5(vethe592bd0) entered disabled state
Jan 12 06:32:14 UnBlue2 kernel: vethe592bd0: entered allmulticast mode
Jan 12 06:32:14 UnBlue2 kernel: vethe592bd0: entered promiscuous mode
Jan 12 06:32:17 UnBlue2 kernel: eth0: renamed from veth32be65a
Jan 12 06:32:17 UnBlue2 kernel: br-f687e37b0ff7: port 5(vethe592bd0) entered blocking state
Jan 12 06:32:17 UnBlue2 kernel: br-f687e37b0ff7: port 5(vethe592bd0) entered forwarding state
Jan 12 06:32:17 UnBlue2 Docker Auto Update: Restarting qbittorrent
Jan 12 06:32:17 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth4f19198) entered blocking state
Jan 12 06:32:17 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth4f19198) entered disabled state
Jan 12 06:32:17 UnBlue2 kernel: veth4f19198: entered allmulticast mode
Jan 12 06:32:17 UnBlue2 kernel: veth4f19198: entered promiscuous mode
Jan 12 06:32:18 UnBlue2 kernel: eth0: renamed from vethbea66d4
Jan 12 06:32:18 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth4f19198) entered blocking state
Jan 12 06:32:18 UnBlue2 kernel: br-f687e37b0ff7: port 23(veth4f19198) entered forwarding state
Jan 12 06:32:18 UnBlue2 Docker Auto Update: Restarting readarr
Jan 12 06:32:18 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth1cc6da4) entered blocking state
Jan 12 06:32:18 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth1cc6da4) entered disabled state
Jan 12 06:32:18 UnBlue2 kernel: veth1cc6da4: entered allmulticast mode
Jan 12 06:32:18 UnBlue2 kernel: veth1cc6da4: entered promiscuous mode
Jan 12 06:32:19 UnBlue2 kernel: eth0: renamed from veth1a70311
Jan 12 06:32:19 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth1cc6da4) entered blocking state
Jan 12 06:32:19 UnBlue2 kernel: br-f687e37b0ff7: port 24(veth1cc6da4) entered forwarding state
Jan 12 06:32:19 UnBlue2 Docker Auto Update: Restarting speakarr
Jan 12 06:32:19 UnBlue2 kernel: br-f687e37b0ff7: port 27(veth3be93b0) entered blocking state
Jan 12 06:32:19 UnBlue2 kernel: br-f687e37b0ff7: port 27(veth3be93b0) entered disabled state
Jan 12 06:32:19 UnBlue2 kernel: veth3be93b0: entered allmulticast mode
Jan 12 06:32:19 UnBlue2 kernel: veth3be93b0: entered promiscuous mode
Jan 12 06:32:20 UnBlue2 kernel: eth0: renamed from veth5f04f20
Jan 12 06:32:20 UnBlue2 kernel: br-f687e37b0ff7: port 27(veth3be93b0) entered blocking state
Jan 12 06:32:20 UnBlue2 kernel: br-f687e37b0ff7: port 27(veth3be93b0) entered forwarding state
Jan 12 06:32:20 UnBlue2 Docker Auto Update: Community Applications Docker Autoupdate finished
Jan 12 06:37:35 UnBlue2 flash_backup: adding task: /usr/local/emhttp/plugins/dynamix.my.servers/scripts/UpdateFlashBackup update
Jan 12 07:37:36 UnBlue2 flash_backup: adding task: /usr/local/emhttp/plugins/dynamix.my.servers/scripts/UpdateFlashBackup update

You can see there that the last entry happened at Jan 12 07:37:36. An user of my Plex confirms that they were using it at 08:25.

Edited January 12Jan 12 by BlueSialia

There's still nothing logged, suggesting more of a hardware problem.

Any idea on how to find that hardware problem? Memtest passes. SMART checks are clean. What else can I do?

Typically RAM, CPU, board or PSU. You posted two different diags; which one is this server, the AMD or Intel?

The problematic server is the Intel one.

Because memtest is only definitive if it finds errors, since you have multiple RAM sticks, you can try using the server with just one, if the same try with the other one, that will basically rule out bad RAM.

It could also be the CPU and the Intel 13/14 gen issue, though it mostly affects the K models, there have been multiple confirmed non-K also bad here in the forum.