Helmonder
-
Posts
2817 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Report Comments posted by Helmonder
-
-
Ok.... did somethin else now..
The not working 10GB was eth0, which was now in a bond with eth1 and 3. eth1 and eth3 are regular lan ports on my motherboard.
- I now swapped eth0 and eth1 so the 10GB is now my 10GB.
- Then I removed the 10Gb eth1 from the bond (connectivity still ok)
- Then I tried to configure the 10GB interface.. AND THE THING GOT AN IP ADDRESS !!
That means that the card itself -is- working on an OS level... This is getting more and more curious...
- Now I moved the eth1 10GB card back into the bond
- Tested connectivity: back on 10GB speeds !
- Did a reboot half expecting it to not work again but alas... it still works !
-
1
-
-
That is what I had... just the 10GB... perfectly working in 6.8.3..
But when I then update to 6.9 there is no network, see
I only made this bond to be able to get 6.9 running... The situation basically is the same only the other nic's in the group do work which is why my server is reachable.. If I now remove the bond that will only result in a server not beiing reachable anymore..
-
Please note this worked fine oit of the box in 6.8.3
-
What do you want me to do with it ? It is installed and shows info but nothing I think I can do something with ?
-
Changed Status to Solved
I did a hail mary..
I added my onboard NICs to my mlx4... After that the update worked and I had networking..
Attached are diagnostics of prior and after this situation..
tower-diagnostics-20210308-1914 - bonded.ziptower-diagnostics-20210308-1932 bonded on 6.9.zip
Now to see if I actually have 10GB , or if I am now running on backup..
-
2 minutes ago, John_M said:
Instead of recreating your USB device from scratch, what happens if you run the appropriate "make-bootable" script for your OS?
I do not have an issue getting the system to boot anymore... Switching to UEFI boot solved that.
-
I solved this by using UEFI boot in stead of Legacy...
-
I did some extra checks, the above might be a bit garbled with info:
1) I have a working, running 6.8.3, I have rebooted just to be sure the usb drive works: it does
2) I run an upgrade to 6.9, this results in a "BOOT ERROR" , before stuff gets loaded so it looks to me like the bootloader is not recognised or something
3) I copy back all of my 6.8.3 files over the 6.9 migrated usb stick, the stick continues to not work and give me a "BOOT ERROR" when booting, I think that shows its not the files., dockers, plugins (also: the issue also occurs with a safe boot)
If I want to make it work again I have to recreate the thumbstick and copy back my 6.8.3 files, then I have a perfectly working system.
This triggered me, if it is to due with the bootloader/bootpartition.. Maybe its linked to my way of booting ? I am using legacy boot..
I changed to UEFI and now I have been able to boot !
However... No network... No GUI, no shares. I was able to use console:
Route-N and IP Route Show output attached.
Both look fine to me...
I grabbed the syslog and also attached that to the post:
I also started the system in safe mode. This does not make a difference, also no network. I also grabbed the syslog fro mthe safe mode:
I am no expert here but I do notice that there -is- a message "eth4: link up", but there is no message for "eth0: link up". Should that not be there ?
-
Well you are not alone... I have the same issue:
The following worked for me (have done it several times trying to get 6.9 to work:
1) Create your USB new with the USB tool (that will clean it totally and make it fresh)
2) Copy the backup you made before upgrading to 6.9 back over it.
3) Reboot
All should work..
Very curious on this boot error thing..
-
On 3/5/2021 at 8:50 PM, limetech said:
Something very odd happening which should be "impossible". I suggest you restore the flash device from backup using the Flash Creator tool. Then try the upgrade again but boot in 'safe mode'.
Did this, same result (see below for my journey here but syslog with this "safe mode" herewith also attached..)
Kind of baffled..
I recopied the 6.8.3 files to the stick and rebooted: all working again..
There is something in my existing flash drive config that breaks this... and Safe Mode does not solve it...
Let me know what you want me to test, not a problem whatsoever !
-
Ok now I am totally puzzled..
I just reimaged my flashdrive with the imaging tool, I put 6.9 on it and just booted with that.. Made no changes to it..
So then there is no boot error (I even tried a couple of times, eveyr time the system boots)
I also get the blue screen asking for how to boot (GUI / No Gui / etc), I then started with the regular unraid, that boots, I even get network (so that card now seems to be working), but when logging on to the system I can enter with "root" only (no password is asked) and my file system is not looking the way it should...
I was able to grab the syslog (syslog_sat.bad), maybe somebody can give me a pointer here ? No idea where to start now..
Just to be sure I also copied my copy op the upgraded usb stick (so in 6.9) over it and rebooted.
That rebooted but the network is gone again (so back where we started).
Next reboot I did was in SAFE mode (so new stick with 6.9, contents of 6.9 copied over, but booted in safe mode). This results in the same situation, no network. syslog is attached as syslog_notok.
-
Ok this is weird... booting seems to not start at all with a "boot error" message... When I recreate my usb stick and copy 6.8.3 over it, then it works again... Tried several times...
I will now recreate my usb stick (with the image tool) and copy over the 6.9 files... curious if that works.. This does appear to be something totally different then the networking that this started out with..
-
This is the diagnostics with the reboot after removing /extra, so still on 6.8.3.
Now rebooting with 6.9..
I'll get back in an hour with the results !
-
@Squid Thanks for pointing that out.... I wasn't even aware I still had that.. I will remove those first as I do not need them anymore and reboot within 6.8.3. Just to be sure. Hope that helps the issue since then its my fault !
@limetech I remade my USB by creating it anew with the usb drive creator and then copying over the contents of my existing.. (with that boot/extra stuff removed). Maybe I need to make a fresh usb in all and only copy the config files that are needed for the basic array.. Have been using unraid for a great while so maybe there is more old sh*t in there..
After that I will try again and post syslog
-
Just tried to do it again while having a bond active on my br0 that also had the native ports of my motherboards in it...
that resulted in a "boot error"
I think I'l wait for the next version
-
Still stable... no errors at all and a very "relaxed" syslog.. No warnings or errors since the 5th..
-
My mistake... I meant the virtio drivers.. Sorry forf the confusion
-
Since the error was referencing KVM and i saw that old libvirt version I have updated it and recreated the libvirt file.. After doing that I saw some old VM's coming back that had not been active, but my correct one (the one that was active) was not in that list... I deleted the wrong ones and re-added (vdisk was ofcourse still there) the missing VM, worked fine.
I have also pinned two threads to the VM (and two other threads to my plex docker).
I have since both actions not seen any page allocation failure messages.. As they appeared to be there on a daily basis and it has now been stable for 48 hours these might have made a difference..
-
Did notice that I was using quite old virtio drivers.. Maybe that could have something to do with it.. Also recreated my libvirt file just to be sure..
-
Will do !
10:03 disabled my VM (one server) and the VM service as a first try.
Since last reboot (This morning) no errors in the log.
Maybe there is a relation to this thread:
-
Bonienl: Not sure since it took the system a week to crash (in the betas and also now) It would be a bit of a hassle to leave the system in safe mode for a week.
I posted to let people know what happened. If the issue does not come back in a week the issues most likely lies in one of the plugins.. I can then bring them back one by one.. Kind of a "safe mode light" scenario. Should the system still crash in a week then I will bite the bullet and let it run in safe mode..
Thanks for your response !
-
On 10/28/2019 at 6:16 PM, Helmonder said:
I upgraded two days ago and have had two crashes since.. All OOM ... Parity checks were running at that time and there were references to cache_dirs .. Unfortunately I have not been able to salvage logs..
System just crashed again.. No webgui, no telnet.. This time no parity check was running.
See attached for a screenprint of my IPMI ..
Since I have IPMI I was able to access the console and have been able to write my syslog to /boot.. It is attached here.
I am seeing (I think) a lot of out of memory type errors.
I am now reverting to the previous release.
-
I upgraded two days ago and have had two crashes since.. All OOM ... Parity checks were running at that time and there were references to cache_dirs .. Unfortunately I have not been able to salvage logs..
-
They seem to be gone since the last update..
Mellanox MT26448 not working on 6.9 (but working great on 6.8.3)
in Stable Releases
Posted
Changed Status to Solved