New install hangs after a few reboots

… and it will not reach the login screen !
The system hangs

A start job is running for /dev/disk/by-uuid  15d401cd-7c17-4874-9cb3-9762f85a5689 (… nolimit)

That is where the system hangs initially.

Next msg :
[ 1315.861064] [T10] memory failure …...recovery action for non structure page pfn: ignored

that’s the spot where it finally hangs

by using ls and blkid commands I obtained uuids for all partitions . Non of the partition uuid correlates to the above figure. (all other ssd where disconnected)

The SSD is a nvme type !

I suspect that there is a hardware fault .

What are your ideas ?

Thanks and best regards

@ozotto what is in the outputs from cat /proc/cmdline and cat /etc/fstab

@malcolmlewis
…sorry about the pic…no cut and past available

@ozotto that’s a dracut shell… sounds like you need to do a fresh install and probably quicker. Is this a single boot setup?

I would run memtest86+for a few hours.

That is a package that you can install and when run grub-efi it will give you another boot option to boot memtest86+. Does not work with grub-bls (yet).

Alternative is to create an USB dongle with the memtest86+ image and run it that way.

1 Like

@malcolmlewis
… it is single boot setup.
I did a number of fresh installs (!) which failed because /boot/efi could not be found.
However , in all cases I did create/specify a FAT partition /boot/efi. …that however was not visible during the next install process.
The system for some reason ignored my entries ???!!!??

@marel
Memtest86 is part of the Tumbleweed-installation submenu.
I already took one DDR4 bar out.
Currently running with the remaining DDR4 bar memtest86
… Pass 44% => zero errors

@marel

One full pass zero errors

Not enough, do it for a few hours. :face_with_monocle:

For me, I built a (still running) desktop machine in 2017 (2018?) that has four Corsair Vengeance sticks (total 128gb).

About 2 years ago, I suddenly saw random “memory issue” notifications.

What I did for a ‘quick test’ (if that’s really something) is I shut the machine down and swapped the DDR sticks in slots 1 and 2.

Then booted up - still random memory issue. So, I swapped the sticks in slot 3 and 4. Also, when swapping, I used a lint-free (microfiber) cloth to swipe the DDR’s contact pins, plus used canned air to blast the DIMM slots. After this last swap, have not had any memory notifications.

1 Like

However … as I read another Reply of yours:

I question whether your issue is really memory-hardware related.

@myswtest

…well, I made a mistake creating /boot/efi !
To format a (any) partition with FAT and mount it /boot/efi does not (!) do the job!

You must create a Partition and select what it is used for => boot/efi , then you format
FAT and mount /boot/efi. (all done within YAST)
By selecting the usage some extra info , so I assume , will be added to the partition.

Anyway , did all that => and it worked … no further problems

but !!! that does not explain my initial problems , it should have worked since the original partition was available.

cheers

Did you reuse partitions / use some tool to partition before install? If the first, did you set the partitioner during install to format the partitions? If the latter, than that could be your issue.
On memtest86+: I have seen machines that only reported errors after letting it run all night, where I had been keeping an eye on it for at least 4 hours. One pass without errors doesn’t mean much.
Also, are the RAM bars the exact same brand/type? I had serious memory issues with a fully new home built machine with 4 DDR4 16GB bars, 2 Crucial, 2 Hynix. Memtesting them separatelyfor 2 x 12h gave zero errors, testing all 4 of them did report errors on 1st pass.

yes ! … so what exactly was the issue ?

No idea yet,I was wondering and trying to debug what was going wrong.

So, to make sure, you now have a working Tumbleweed?

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.