3.8.5 cant acces foris after update - router dont respond after factory reset

Software SW help
1

Hi I have this issue after update to 3.8.5
I cant acces Foris interface on one of the my omnias

error:

any tips?

also I try rollback via schnapps but when I use command for list of backups I get this erros


At this moment I dont have acces to the router for factory reset, is there any chance to solve it remotely?

I think something bad happend with btrfs file system is read-only

LXC containers are offline, Pihole,nextcloud ,etc,
whole network is down beacuse Pihole is a DNS, please help if it possible

2

Please di not reboot it yet and send us complete logs and kernel output (file /vat/log/messages and output of command dmesg). We might catch what happened. But yes it seems to me that you have corrupted filesystem. So please also apppend output of command mount.

3

Hello cynerd, I have sad news, my friend trying factory reset this omnia in night - so logs are gone

now problem is router is ´´bricked´´ after factory reset, led for wan and one of LAN port blinks but we Cant acces omnia on 192.168.1.1

4

I dont know how it possible but router is back with same issues - 500 internal server error
I wait little bit for logs, this data router collect:
dmesg log:

var log messages log here:

opkg list-installed:

fount this about BTRFS

2017-11-11T15:06:25+01:00 warning kernel: [ 55.837508] BTRFS warning (device mmcblk0p1): mmcblk0p1 checksum verify failed on 365981696 wanted 53064BF7 found CDDEA918 level 0
2017-11-11T15:06:25+01:00 info kernel: [ 55.837529] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 0
2017-11-11T15:06:25+01:00 warning kernel: [ 55.841301] BTRFS warning (device mmcblk0p1): mmcblk0p1 checksum verify failed on 365981696 wanted 53064BF7 found 867232BA level 0
2017-11-11T15:06:25+01:00 info kernel: [ 55.841322] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 4096
2017-11-11T15:06:25+01:00 warning kernel: [ 55.845270] BTRFS warning (device mmcblk0p1): mmcblk0p1 checksum verify failed on 365981696 wanted 53064BF7 found 2316E253 level 0
2017-11-11T15:06:25+01:00 info kernel: [ 55.845296] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 8192
2017-11-11T15:06:25+01:00 warning kernel: [ 55.849072] BTRFS warning (device mmcblk0p1): mmcblk0p1 checksum verify failed on 365981696 wanted 53064BF7 found 719E960 level 0
2017-11-11T15:06:25+01:00 info kernel: [ 55.849095] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 12288
2017-11-11T15:06:25+01:00 warning kernel: [ 55.849870] BTRFS warning (device mmcblk0p1): csum failed ino 8387 off 0 csum 1547075464 expected csum 0
2017-11-11T15:06:25+01:00 warning kernel: [ 55.849904] BTRFS warning (device mmcblk0p1): csum failed ino 8387 off 4096 csum 467442729 expected csum 0
2017-11-11T15:06:25+01:00 warning kernel: [ 55.849934] BTRFS warning (device mmcblk0p1): csum failed ino 8387 off 8192 csum 1027143069 expected csum 0
2017-11-11T15:06:25+01:00 warning kernel: [ 55.849964] BTRFS warning (device mmcblk0p1): csum failed ino 8387 off 12288 csum 1205327285 expected csum 0
2017-11-11T15:06:25+01:00 info kernel: [ 55.853788] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 16384
2017-11-11T15:06:25+01:00 info kernel: [ 55.857594] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 20480
2017-11-11T15:06:25+01:00 info kernel: [ 55.861419] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 24576
2017-11-11T15:06:25+01:00 info kernel: [ 55.865277] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 28672
2017-11-11T15:06:25+01:00 info kernel: [ 55.869096] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 32768
2017-11-11T15:06:25+01:00 info kernel: [ 55.872949] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 36864
2017-11-11T15:06:25+01:00 info kernel: [ 55.876880] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 40960
2017-11-11T15:06:25+01:00 info kernel: [ 55.880690] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 45056
2017-11-11T15:06:25+01:00 info kernel: [ 55.884542] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 49152
2017-11-11T15:06:25+01:00 info kernel: [ 55.888249] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 53248
2017-11-11T15:06:25+01:00 info kernel: [ 55.892050] BTRFS info (device mmcblk0p1): no csum found for inode 8387 start 57344

5

Yep. Corrupted filesystem. You have to do four leds reflash from usb drive. That should help. But please if you encounter it again, it’s not first time someone is reporting such thing, I would be glad to receive kernel log, syslog messages and mountpoints to see how it happened. We didn’t encountered it so we really don’t know how it’s happening.

Just to verify something, you have lxc containers on internal memory?

6

yes in internal memory

if you need some others log please tell me what exactly you need, in next 2 day I can reflash omnia from usb drive, Or my friend do it today

7

Well I clearly see that it is because btrfs has unrecoverable index. What I need is log from point when it happened. You can’t provide that to me because you have already rebooted router. So feel free to reflash it.

I would suggest you to have lxc containers on external drive as I am currently suspecting from these errors software running in containers. Specially journald and other logging or database software. It is probably an error in BTRFS it self but it seems to be triggered by greater amount of writes. But that is no way final and as I had not managed to reproduce it, so I can’t say for sure that this is what it is.

1 Like
8

Is 03:24 AM here and I trying reflash this omnia, but I get stucked again and again, I think this omnia is death piece of wonderful hardware

now this is a problem:
I format usb to FAT, copy medkit on it, when I push reset for 4 leds, all leds blinks for green loading and now all leds are red for 15 min and still not change

I make something wrong? I never reflash turis omnia before sorry for noob question

9

I’m in similar situation. I’ve started 4 LED reset and router is now stuck in all red state for 8+ hours. Output on serial port is:

mmc0: Got data interrupt 0x00100000 even though no data operation was in progress.

Is reflash still in progress or is it safe to restart the router now?

10

What kind of FAT? If I am not mistaken then only FAT32 works. Be sure that you don’t have it formated as vFAT or exFAT. Also it has to be formated with partition table. To be sure that you have it in linux that means that you have formated something like /dev/sdx1 and not /dev/sdx. Please check all of this and get back to me. But true is if they are red for long time then it should be fine and it should be working.

So if router has no progress the if you have serial cable at hand then please use it and see what is happening.

11

Well if it’s more then few minutes (about ten) then there is probably problem somewhere. When you have serial cable connected then please reboot router back to reflash and record whole process from beginning.

12
13

@detto this seems to be broken flash. Please contact our support if you have bought router from campain or conntact reseller you bought it from to request warranty repair.

Same question here. Have you had any lxc container on flash or have you been overall writing to flash with some program?

14

Thank you for confirmation.
Yes I had LXC running on flash.

15

I follow this https://www.turris.cz/doc/en/howto/omnia_factory_reset and format USB for FAT as guide say

I try FAT32 then I can format this usb in windows ?

16

is not possible to reflash this omnia , atm I dont have serial cable here

I try conntact reseller and send router for warranty repair …

We cant use LXC containers in internal memory?

17

It should be possible. Unless of course it’s permanent flash failure. I can’t tell without serial console. But more probable is that your flashdrive is just invalidly formated or that file on it has invalid name. Please once again check that you have usb drive with partition table and with first partition containing FAT or ext with medkit. Also to be sure you can disconnect any external drives (or even internal ones if you have sata).

You can use lxc in internal memory. But problem is that software in standard distribution is not configured for flash usage. You can easily run lxc on internal storage but for example enabling verbose logging will result to weared out flash. So there is nothing problematic with lxc it self but woth it’s possible usage. But the same thing can happen not only with lxc but also for example with mysql, we have possibility to have mariadb on router and missuse of that database can lead to flash damage. I have some warning tool in my backlog but until then I would suggest you to use lxc where ever possible and store /srv on external drive.

1 Like