Bug #10702
closed
Todays Snapshot System does crash and does not start any more
0%
Description
Hello,
Two hours ago I did upgrade to latest snapshot, during the related reboot the system never came back.
I do not if there is any way of going back to the previous kernel and config (idea there should be an A- and B-side).
What ever:
- this is critical, I did take precausions (backups install files on usb, had to do it before :( etc
that made it posible to install an older version in relatively short time (1,5 hour!!)
There are two probably related messages in the boot-logging starting just after "performance statistics done"
1) fopen(): filename cannot be empty in /etc/inc/notices.inc on line 75
2) cannot use scalar value as an array in /etc/inc/config.lib.inc on line 383
Files
| 20200625_143747(0)_capture_resized_1.jpg (218 KB) 20200625_143747(0)_capture_resized_1.jpg | Here the problem starts | ||
| 20200625_144209_resized.jpg (773 KB) 20200625_144209_resized.jpg | There the resulting endless loop | ||
| textdump.tar(2).0 (92 KB) textdump.tar(2).0 | |||
| info(2).0 (418 Bytes) info(2).0 |
_capture_resized_1.jpg){kind=link}
Updated by Jim Pingle almost 4 years ago
- Status changed from New to Not a Bug
- Priority changed from Urgent to Normal
- Target version deleted (
2.5.0)
No problems here on the latest snapshot. That seems more like you have a local failure (perhaps a storage problem or corrupted filesystem, or hardware)
This site is not for support or diagnostic discussion.
For assistance in solving problems, please post on the Netgate Forum or the pfSense Subreddit .
See Reporting Issues with pfSense Software for more information.
Updated by Louis B almost 4 years ago
- File 20200625_143747(0)_capture_resized_1.jpg 20200625_143747(0)_capture_resized_1.jpg added
- File 20200625_144209_resized.jpg 20200625_144209_resized.jpg added
Hereby two fotos captured via mobile using the "KVM-switch". Perhaps helpfull.
Louis
Updated by Jim Pingle almost 4 years ago
See my previous comment. That is most likely a problem with your installation, not a bug in the snapshots. Main suspect is your storage (SSD/HDD/etc) or a corrupt filesystem, errors are likely from a missing/corrupt config.xml among other problems.
Updated by Louis B almost 4 years ago
Jim,
I did a new install using the today 650 snapshot, since the update had failed.
Then I installed the config from earlier today, -- not good ---.
One of the issues the packages, did not reinstall automaticly
When I tryed to do it by hand first stopping the automatic process, that was not possible another proces did lock the install .....
Looking into the config I noticed that there was a left over from an earlier ffr-install (hum ... hum ...)
Tryed again -- not good -- again
However pfSense was running and I did generate a config without package info
And did load that
it worked ....
however the package config info was gone ....
but the system did install the packages .... strange
I also did see it the other way around in the past
- remove a package
- install it again
- and wonder ....... the configuration data was still there
This to give you background info
..... however the issues described above ...... are more than good for a lower prio bug to fix IMHO
Louis
Updated by Jim Pingle almost 4 years ago
You should post these kinds of things on the forum and not here. That is the proper place to discuss and diagnose what might have happened and then once a real problem is identified (unique, unknown, reproducible) an issue can be opened with details about what happened and how to reproduce it and so on.
I've tested numerous VMs on todays snapshot and had no problems, it must be something in your methodology or environment.
It's also possible you are hitting other known issues like #10610 which is already being investigated.
No matter what, however, your first instinct should not be to assume every problem you have is a bug to be entered in this issue tracker. Start with the forum.
Updated by Louis B almost 4 years ago
Jim
the reason to post it here, is that I was very afraid that other people would become in trouble
And depending on peoples config that is perhaps still true.
For that reason, you should see the message as soon as possible IMHO, and that is not at all guaranteed when using the forum
Louis
Updated by Jim Pingle almost 4 years ago
We constantly monitor the forum, that is not a concern.
Updated by Louis B almost 4 years ago
- File textdump.tar(2).0 textdump.tar(2).0 added
- File info(2).0 info(2).0 added
Jim,
I can easely reproduce this issue on a running system:
- disable on of the interfaces
- stop pimd (e.g. pimd q, but I think stopping it in another way is ok as well) enable the interface again
==> that was the end of the pleasure
- in case the system is still running start PIMD but I do not think you will come that far
On the KVM-switch I see the same messages as described before
I did attache a crash dump, which I think is related to the same issue. I do have an exact crash dump because I did need to format the disk and reinstall pfsense again
By the way I did use
2.5.0-DEVELOPMENT (amd64)
built on Fri Jun 26 01:03:28 EDT 2020
FreeBSD 12.1-STABLE
this time
Louis
Updated by Louis B almost 4 years ago
Jim,
It is not related to PIMD. In the past week I had many crashes related to swithing interfaces on or off yes or now combined with a reboot. Because I initionally did expect a direct relation with PIMD, I removed all packages including PIMD. Even then I see the problem.
Louis
Updated by Chris Palmer almost 4 years ago
Louis- There are no problems with the snapshots on my end either. I try to test every release that comes out. Please post these issues in the forum so others will see and can decided to test or not.