August 11, 20232 yr Hi, Please help if you can. I'm experiencing frequent server crashes, I suspect that it might be my cache drive (swapped SATA cable for a port on my HBA) as that has been giving errors in the past week. I can't see what has caused the crash this time. I have recently upgraded to 6.12.3 and installed new Mobo/CPU/RAM. Regretting doing all the upgrades at the same time now. Attached diagnostics and additional log files tower-diagnostics-20230811-1531.zip syslog-10.10.22.125.log
August 11, 20232 yr Community Expert Not seeing anything relevant logged, suggesting more a hardware issue, do you know the day/time it last crashed?
August 11, 20232 yr Author Latest crash was around 19:54:43. I've attached a new diagnostics and syslog that is saved to the array. Below is the logs around the latest crash, I removed a bunch of `ACPI action...` logs that were flooding the logs. They were caused everytime I pressed an arrow key on the keyboard. I attempted to connect to the console after the crash to download diags / logs but no luck, so I don't think the logs are as complete as they could be.. Aug 11 18:44:45 Tower webGUI: Successful login user root from 10.10.22.110 Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered blocking state Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered disabled state Aug 11 18:46:24 Tower kernel: device vethe89368e entered promiscuous mode Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered blocking state Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered forwarding state Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered disabled state Aug 11 18:46:24 Tower kernel: eth0: renamed from veth7fadde6 Aug 11 18:46:24 Tower kernel: IPv6: ADDRCONF(NETDEV_CHANGE): vethe89368e: link becomes ready Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered blocking state Aug 11 18:46:24 Tower kernel: docker0: port 8(vethe89368e) entered forwarding state Aug 11 18:55:15 Tower kernel: mpt3sas 0000:01:00.0: invalid VPD tag 0x00 (size 0) at offset 0; assume missing optional EEPROM Aug 11 18:52:51 Tower root: ACPI action up is not defined Aug 11 18:55:27 Tower root: ACPI action down is not defined ... ... Aug 11 19:02:23 Tower root: ACPI action right is not defined Aug 11 19:12:40 Tower emhttpd: spinning down /dev/sde Aug 11 19:12:42 Tower emhttpd: spinning down /dev/sdh Aug 11 19:12:42 Tower emhttpd: spinning down /dev/sdg Aug 11 19:12:42 Tower emhttpd: spinning down /dev/sdd Aug 11 19:12:42 Tower emhttpd: spinning down /dev/sdf Aug 11 19:54:43 Tower emhttpd: spinning down /dev/sdb Aug 11 21:14:32 Tower root: Delaying execution of fix common problems scan for 10 minutes Aug 11 21:14:32 Tower unassigned.devices: Mounting 'Auto Mount' Devices... Aug 11 21:14:32 Tower emhttpd: Starting services... tower-diagnostics-20230811-2140.zip syslog-10.10.22.125.log
August 11, 20232 yr Author I am seeing this in the logs that indicates an error with my HBA. I swapped the slots that my HBA is connected to and the error followed the device. Line 577: Aug 11 10:53:36 Tower kernel: mpt3sas 0000:02:00.0: invalid VPD tag 0x00 (size 0) at offset 0; assume missing optional EEPROM Line 1313: Aug 11 18:55:15 Tower kernel: mpt3sas 0000:01:00.0: invalid VPD tag 0x00 (size 0) at offset 0; assume missing optional EEPROM
August 11, 20232 yr Author Ok red herring then, I'll ignore it. Would it be worth running a memory test to rule that out? I'm getting sick of rebooting the server so won't mind the downtime now! Edited August 11, 20232 yr by jdndm
August 11, 20232 yr Author I think I just managed to catch it crashing from the console. Apologies for the crappy photo. I rushed to snap it. After this photo was taken, the server rebooted, the HBA loaded and discovered the attached drives. The BIOS failed to load after that, so nothing else booted. I'm stuck on a blank screen with a flashing cursor.
August 11, 20232 yr Community Expert 56 minutes ago, jdndm said: I think I just managed to catch it crashing from the console. Console shows a shutdown being initiated, this can be done by the OS or a plugin, or a quick power button press. 57 minutes ago, jdndm said: I'm stuck on a blank screen with a flashing cursor. Check that the flash drive is bootable, or redo it.
August 11, 20232 yr Author Well since that last shutdown I have been unable to get my system to boot up at all. I now have a "USB device overcurrent warning" and can't even load the BIOS. I've unplugged everything except a single stick of ram, CPU and 1 fan but still no luck. I'm suspect the motherboard has failed. Thanks for your help, but I'm going to have to leave it here for now and hope I have better luck with some different hardware.
June 29, 20242 yr Well i wanted t add my 2 cents here on my crashing on unraid, pst update so not replaced to macvlan. I replaced my USB with a new one. and it didnt crash all day, started my dockers and it crash almost immediately so I started looking at ,NVME, motherboard, ram and CPU. I ordered a new motherboard currently h370, going to a z390Pro and while waiting I starting doing some digging I downloaded the memtest86 image and flashed to a new stick. Crash 41% two times. moved ram around continued crashing. Tested with one stick no longer crashing and the array still runs (16gb) single stick. I'm going to test again on the other one in the same slot next, that'll let me know if its MB or ram as the cause.
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.