Startup problems after wrong antiX shutdown

[Member marcelocripe]

Title in “en”: Problems starting up after wrong antiX shutdown.

After a wrong antiX shutdown (ISO “antiX-19-legacy-bet2_386-base.iso”) due to an interruption in the power supply, problems with startup occur, with error messages during the loading of antiX in text mode:

https://ibb.co/4NdrNTQ

When the graphics mode is loaded and Slim is displayed, entering the username “felipe” and the password to access the session with your default account, the message: “Falied to execute login command” was displayed:

https://ibb.co/j4cqMHh

We were able to log in with the “root” account that loaded the desktop with the Fluxbox window manager.

Returning to the Slim screen, I typed the command “exit” which allowed me to access the user “felipe” in text mode.

https://ibb.co/y660Xqf

I tried to help @felipecardope solve this problem, but I couldn’t. I still don’t have enough domain on the antiX Linux operating system and at the same domain level as I do on the Windows operating system. So, I ask for help from the most experienced.

What can we do in antiX when wrong shutdowns occur due to a power outage and the operating system no longer starts in graphical mode?

We already thank anyone who can help us.

marcelocripe
(Original text in Brazilian Portuguese)

———-

Título original em pt-BR: Problemas na inicialização após desligamento errado do antiX.

Após um desligamento errado do antiX (ISO “antiX-19-legacy-bet2_386-base.iso”) devido a interrupção do fornecimento de energia elétrica, ocorrem problemas na inicialização, com mensagens de erro durante o carregamento do antiX no modo texto:

(Ver a URL da primeira imagem no texto em inglês)

Quando o modo gráfico é carregado e é exibido o Slim, inserindo o nome de usuário “felipe” e a senha para acessar a sessão com a sua conta padrão, a mensagem: “Falied to execute login command” foi exibida:

(Ver a URL da segunda imagem no texto em inglês)

Nós conseguimos iniciar a sessão com a conta de “root” que carregou a área de trabalho com o gerenciador de janelas Fluxbox.

Voltando para a tela do Slim, digitei o comando “exit” que me permitiu acessar o usuário “felipe” em modo texto.

(Ver a URL da terceira imagem no texto em inglês)

Eu tentei ajudar o @felipecardope a resolver este problema, mas eu não consegui. Eu ainda não possuo domínio no sistema operacional antiX Linux suficiente e no mesmo nível de domínio que possuo no sistema operacional Windows. Por isso, eu peço ajuda aos mais experientes.

O que podemos fazer no antiX quando ocorrem desligamentos errados devido a interrupção do fornecimento de energia elétrica e o sistema operacional não inicia mais no modo gráfico?

Desde já agradecemos a quem puder nos ajudar.

marcelocripe
(Texto original em Português do Brasil)

[Member Robin]

Helpful

Up

0

::

Hello Marcelo,

what you see on the third screenshot is, you have to deal with an corrupted (inconsistent) file system, which is an expectable consequence of power supply failure.

The screen output also gives you a hint what to do next:

$ cat /var/log/fsck/checkfs

and post its output here.

Moreover you could try to follow the second advice from screen output: “call fsck without any arguments”

so simply type

fsck <partition>

where <partition> is to be replaced by something like e.g “/dev/sda1” or “/dev/sdb2”, and let us know the output. Best would be to perform this check after having booted from another USB stick, since the file system to be repaired should better not be mounted. In this case you will have to find out the correct device name of the damaged partition before.

lsblk

will give you a first overview.

if fsck was able to fix the inconsistency try to reboot after fsck did all its work.

Is this an hdd install or a live system from usb?

good luck!

Robin

• This reply was modified 2 years, 1 month ago by Robin.
• This reply was modified 2 years, 1 month ago by Robin.
• This reply was modified 2 years, 1 month ago by Robin.

Windows is like a submarine. Open a window and serious problems will start.

[Member ModdIt]

Helpful

Up

0

::

Hi Marcelo, felipe,
I hope you have a backup of your system.
If not:
First job always make a backup before trying to fix the filesystem. I have seen
a computer where fsck fixed it to an unrecoverable condition.

Many times files can be rescued although corrupted, they have weird characters or
symbols but are (almost) always editable.

If you always install with a separate home and have a backup of that as well as a
bootable copy of your system, you are pretty safe.

If doing fsck to try to fix the file system does not work and you can read files
on home, you probably have damaged configuration files.

Easy/safest way is create a new user from root account, ensure you can login to that user.
Copy over your files and work, either from the non working user, alternative from a backup.

Delete old user afterwards. The base system is running so you should not need a reinstall.

In Manila and Porto Galera Power outages came often, Got used to above.

[Member marcelocripe]

Helpful

Up

0

::

Hi Robin and ModdIt,

Thank you for your quick responses.

In a little while I will go to Felipe, this time I will be prepared with a USB device with antiX, to apply his explanations with him. Sorry, I forgot to link topic that I described the main steps of installing his computer and with photos that he helped me to register.

Is this an hdd install or a live system from usb?

After following some topics on this forum, I understood the importance of separating into different partitions, one partition for the operating system and another for storing user data and settings. At installation the system – “root” is on the sda1 partition and “home” is on the sda2 partition of the hard drive.

As soon as possible I will post the results of the commands.

So, does it mean that it may be necessary to install antiX again on the hard drive?

marcelocripe
(Original text in Brazilian Portuguese)

———-

Olá Robin e ModdIt,

Obrigado por suas repostas tão rápidas.

Daqui a pouco eu irei até o Felipe, desta vez estarei preparado com um dispositivo USB com o antiX, para aplicar junto com ele as suas explicações. Desculpem-me, acabei me esquecendo de vincular o tópico que descrevi as principais etapas da instalação do computador dele e com fotos que ele me ajudou a registrar.

Is this an hdd install or a live system from usb?

Depois de acompanhar alguns tópicos deste fórum, eu entendi a importância de separar em partições diferentes, uma partição para o sistema operacional e outra para armazenar os dados e configurações do usuário. Na instalação o sistema – “root” está na partição sda1 e “home” está na partição sda2 do disco rígido.

Assim que possível eu postarei os resultados dos comandos.

Então, significa que poderá ser necessário fazer a instalação novamente do antiX no disco rígido?

marcelocripe
(Texto original em Português do Brasil)

[Moderator Brian Masinick]

Helpful

Up

0

::

You will only have to reinstall the software if the file system check and repair step (fsck) fails and the system is corrupt.

If only the data is corrupt, use the backups to get back what you can.

If the system still won’t boot after running fsck from the USB, then you may have to reinstall or replace with your backup copy.

--
Brian Masinick

[Member marcelocripe]

Helpful

Up

0

::

Hi Robin,

When the damaged antiX loaded, it reached the Slim screen, typed “exit” and pressed “Enter” to access the text mode.

https://ibb.co/fdn6cVL

I accessed the user “felipe” and applied the command ”$ cat /var/log/fsck/checkfs” and “$ fsck /dev/sda1”:

https://ibb.co/YDC8TLj

With the root user:

https://ibb.co/jf44HHm

Then I applied the command “$ lsblk”:

https://ibb.co/wWKyMWy

Using antiX through the USB executable (Live USB), in the folder “/media/rootantiX19/var/log/fsck” I was able to find the file “checkfs” and copied its contents.

https://ibb.co/6wVkxBF

Log of fsck -C -M -A -a
Sat Apr 10 15:16:26 2021

fsck from util-linux 2.33.1
homeantiX contains a file system with errors, check forced.
homeantiX: Inodes that were part of a corrupted orphan linked list found.

homeantiX: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
fsck exited with status code 4

Sat Apr 10 15:16:27 2021
—————-

I was building the text and during the process, the power supply was interrupted again, which made me lose everything, because I was using antiX via the USB executable (Live USB), I still don’t know how to use data persistence .

I hope I was able to send you all the information you requested. Recovering a damaged antiX is something I want and have to learn, as interruptions in the supply of electricity are frequent, especially after the privatization of the distribution of electricity, the quality has only worsened … The stabilizer manages to maintain for a few seconds, but the ideal would be to have a UPS, but it is very expensive, they are the best devices to protect a computer.

marcelocripe
(Original text in Brazilian Portuguese)

———-

Olá Robin,

Quando o antiX danificado carregou, chegou até a tela do Slim, digitei “exit” e pressionei o “Enter” para acessar o modo texto.

(URL da imagem no texto traduzido em idioma inglês)

Acessei o usuário “felipe” apliquei o comando ”$ cat /var/log/fsck/checkfs” e “$ fsck /dev/sda1”:

(URL da imagem no texto traduzido em idioma inglês)

Com o usuário root:

(URL da imagem no texto traduzido em idioma inglês)

Depois eu apliquei o comando ”$ lsblk”:

(URL da imagem no texto traduzido em idioma inglês)

Utilizando antiX através do USB executável (Live USB), na pasta “/media/rootantiX19/var/log/fsck” eu consegui encontrar o arquivo “checkfs” e copiei o seu conteúdo.

(URL da imagem no texto traduzido em idioma inglês)

Log of fsck -C -M -A -a
Sat Apr 10 15:16:26 2021

fsck from util-linux 2.33.1
homeantiX contains a file system with errors, check forced.
homeantiX: Inodes that were part of a corrupted orphan linked list found.

homeantiX: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
fsck exited with status code 4

Sat Apr 10 15:16:27 2021
—————-

Eu estava construindo o texto e durante o processo, o fornecimento de energia elétrica foi interrompido novamente, o que me fez perder tudo, pois eu estava utilizando o antiX através do USB executável (Live USB), eu ainda não sei utilizar a persistência de dados.

Eu espero ter conseguido te enviar todas as informações que você me solicitou. Recuperar um antiX danificado é algo que quero e tenho que aprender, pois interrupções de fornecimento de energia elétrica são frequentes, principalmente após as privatizações da distribuição de energia elétrica, a qualidade só piorou… O estabilizador consegue manter por alguns segundos, mas o ideal seria ter nobreak, mas é muito caro, são os melhores dispositivos para proteger um computador.

marcelocripe
(Texto original em Português do Brasil)

[Anonymous]

Helpful

Up

0

::

I was building the text and during the process, the power supply was interrupted again, which made me lose everything, because I was using antiX via the USB executable (Live USB), I still don’t know how to use data persistence .

From the LegacyBIOS boot menu, press “F5” and select root_persist
(other choices, other forms of persistence are offered, I am coaching only this particular option)

As a one-time setup, during the boot process, a script will walk you through the steps:
You will choose new passwords for “demo” and for “root”.
You will specify the initial size (capacity) of the persistence storage file.
(“initial” meaning that you are able to resize it larger during later sessions, on-demand)
You will specify a “mode” ~~ Auto, Semi-auto, Manual.
Because semi-automatic provides ability to perform on-demand saves, repeatedly, throughout a livesession… I recommend (and I use) semi-automatic.

[Anonymous]

Helpful

Up

0

::

What can we do in antiX when wrong shutdowns occur due to a power outage and the operating system no longer starts in graphical mode?

In a followup post, I will explain why I think this is “the wrong question to ask”.

Recovering a damaged antiX [filesystem] is something I want and have to learn

This is a generic problem, it is not specific to antiX.
The tools available, and the techniques involved, are also generic.
If you do learn this skill, you will be able to apply it (forensic recovery) across a variety of filesystems and operating systems.

Following the unsuccessful manual fsck (which was advised onscreen)
an IMMEDIATE clean shutdown should be attempted.
In order to preserve the integrity of the data, do not repeatedly boot the drive and poke around. Instead, boot from another operating system and either use the “dd” command to create a safekeeping copy of the content to another device… or (less ideal) mount the affected partition read-only in order to browse and assess the degree of damage to its content.

At this point (multiple lookaround reboots?), the affected (“orphaned”) data blocks of felipe’s “homeantiX” partition may have now been overwritten by subsequently-written files (.Xession-errors, et al) ~~ if the affected partition had been the boot partition, freshly-written bootlog(s) and syslogs and Xorg.log would similarly threaten the likelihood of recovering any orphaned blocks.

homeantiX contains a file system with errors, check forced.
homeantiX: Inodes that were part of a corrupted orphan linked list found.

homeantiX: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)

If the filesystem is ext4, and journaling is enabled,
the “dumpe2fs” and “debugfs” command (ubiquitous across all distributions) plus a hexeditor…
along with the sibling “photorec” and “testdisk” utilities (debian packagename “testdisk”)
in most cases, these will enable a skilled tech to perform a successful recovery.

[Anonymous]

Helpful

Up

0

::

After following some topics on this forum, I understood the importance of separating into different partitions, one partition for the operating system and another for storing user data and settings. At installation the system – “root” is on the sda1 partition and “home” is on the sda2 partition of the hard drive.

Well… somehow, I do not share that “understanding”.
Compared to “separate partitions”, placing $HOME on a separate physical device seems like a more effective way to mitigate risk.

What can we do in antiX when wrong shutdowns occur due to a power outage and the operating system no longer starts in graphical mode?

Is this “the wrong question” to ask?
Probably a better approach would be to ask/investigate How might we preclude future similar problems?

my recommended solution, toward avoiding “power outage and the operating system no longer starts…”
is
liveboot, dynamic root persistence, semi-automatic save
plus, if feasible (and here, my attention is toward storms/lightning)
choose the “toram” liveboot option ~~ detaching the liveUSB device until/unless a persist-save operation is warranted.

^—- plus one more detail (explicit “data=journal” ext4 mount option, but this merits a separate explanatory post)

______________________________________

(using self as an example)
-=-
According to debugfs (and my conky on this machine), here are a couple of stats/details:
1) latest remaster (rootfs was “born on”): 5 Sept 2020
2) To date, since 5 Sept 2020, a cumulative total of (only!) 463MB has been written to the rootfs.
3) trivia: the content of that roofts file presents as (only!) 16 fragments

The bulk of those writes are attributable to system updates (package upgrades)…
yet (however) I am not utilizing a separate “$HOME”.
Frequent iso snapshots, offloaded to LANserver for safekeeping, cover the task of “full system backup” and re-attaching (or reconnecting to, via LAN) external data volumes… this has enabled me to boot from whichever machine I choose and/or workaround any equipment downtime.

six vs half-dozen?
For others, using a separate $HOME may provide a fine experience.
I’m not knocking it, just explaining a “works well for me” alternative.

[Member Robin]

Helpful

Up

0

::

Hi Marcelo,
what skidoo wrote above is absolutely true. (But it will be probably difficult for you to follow some of his instructions)

In order to make some handy steps from it you can easily follow, even after translation to brasilian Portuguese language, I’ll try my best to give you something like a cooking recipe for this:

From the first new screenshot you can learn the damaged partition is /dev/sda1 obviously.
So, keep in mind what I said in my first reply, and skidoo also demanded you to do:

1.) Boot the System from your Live USB device in order to fix anything on the damaged system.

2.) make sure /dev/sda2 is mounted somewhere (e.g. /media/_sda2) and you have enough space left on sda2.
(from your screenshots you can see, sda1 is 20GB, so you need at least 20GB free space on sda2, which has 52GB total)

3.) Make an dd copy of the /dev/sda1 partition for safety, just as skidoo suggested already.
make sure the target path /media/_sda2 exists (or adapt the command below to where you find sda2 being mounted to.)
Then execute from a terminalwindow (Roxterm) the command
dd if=/dev/sda1 of=/media/_sda2/partition_1.img
This will take some time probably, you will not see any progress.
But you can check anyway from time to time :
Open a second terminalwindow (Roxterm) and enter in this _second_ terminal window:
kill -USR1 $(pidof dd)
which will show you some information about the progress of your copy. You can repeat this command from time to time in order do see the progress.

Once finished this copy, you can safely try to fix the first partition:
(until now fsck was never executed correctly, due to the fact the filesystem was mounted, as you can see from your screenshots)
So make sure sda1 is umounted, then start fsck :

sudo fsck /dev/sda1

If you are lucky it will be able to recover your filesystem.
It is absolutely necessary to perform all this from a live system, and the partition is not to be mounted.

@skidoo: can you verify from his screenshots it is really /dev/sda1 which is damaged?
Neither the posted content of /var/log/fsck/checkfs nor the output on screen does contain any visible hint for me, since in the fsck command which actually failed due to mounted file system, they’ve entered explicitly /dev/sda1. Is there any possibility it is /dev/sda2 instead, which is damaged? How could he check this safely before copying to /dev/sda2 ?

@marcelocripe: the mail you’ve received this morning from me contains also some hints how to make your and felipes system more resistant against power supply failure, utilising simply a rotation of two or three antiX live sticks, one or two of them being the reserve of recent (when two sticks) or recent and before update version of your system (when three sticks). This way you will always be able to recover your system within 5 minutes maximum: just shut down, replace the stick, reboot. That’s all. Than immediately recreate the damaged system stick from the reserved one using the antiX tools in order to be safe again for next power supply failure.

• This reply was modified 2 years ago by Robin.
• This reply was modified 2 years ago by Robin. Reason: wordpress censored most of the text originally
• This reply was modified 2 years ago by Robin.

Windows is like a submarine. Open a window and serious problems will start.

[Forum Admin Dave]

Helpful

Up

0

::

I do not often get the fsck message, even when loosing power. I would check the hard drive to see if it is going bad.
smartct -a /dev/sda

When maintaining the installs (versus reinstalling after corruption), after running fsck to check and fix the file system you can check (most of ?) the system files with debsums.
apt update && apt install debsums
dpkg -l > packages.list
open packages.list (nano/cat/less/more) and read the package names into debsums
debsums app-select-antix

It should output if every file matches the checksum of what is included in the package. Note that if you modified a configuration file it will show up as not matching. This does not mean it is corrupt but take notes of them to verify. Afterward you could use a system monitoring program for key files and your configuration files to see if they change from what you expect. A separate home partition is always handy because that can save you from monitoring the files therein. I like to use rsync to copy the contents off the disk to another disk (usb/network). You may want to do disk images every time the system updates as Robin shows with dd, specifically of the beginning sectors/mbr and root partition if running separate partitions (as in tell dd to stop before copying the home partition as this in my experience take the most time). That can make it easier to restore if complete corruption occurs.

Arguably this is all more difficult than backing up your home data, doing a snapshot, and reinstalling the snapshot when corrupt.

Computers are like air conditioners. They work fine until you start opening Windows. ~Author Unknown

[Member marcelocripe]

Helpful

Up

0

::

I thank everyone for the explanations and the teachings.

Tomorrow, after work I intend to go to Felipe, if he authorizes me, I will take the computer to my house and I will continue trying to carry out each of the orientations of each one of you.

Grateful.

marcelocripe
(Original text in Brazilian Portuguese)

———-

Eu agradeço a todos pelas explicações e pelos ensinamentos.

Amanhã, depois do trabalho eu pretendo ir até o Felipe, se ele me autorizar, eu levo o computador para a minha casa e continuarei tentando realizar cada uma das orientações de cada um de vocês.

Grato.

marcelocripe
(Texto original em Português do Brasil)

[Anonymous]

Helpful

Up

0

::

>>> screenshot

When I typed that, I pasted the fsck output (also seen in the screenshot)
“homeantiX contains a file system with errors, check forced.”
marcelocripe can determine whether or not that matches the partition label of sda1

something like a cooking recipe for this

Oh, my point of view was
“the fish got away this time”
or
“the horse has escaped, too late to fuss with fixing the stable door”

If marcelocripe does intend to invest effort into recovering from “this time, this event”, the “diffoscope” (apt install diffoscope) program could be used to compare against the files present on a known-good (aka freshly installed, the dist-updated) filesystem.

[Anonymous]

Helpful

Up

0

::

Whoops, noticed a typo from a few posts back:

According to debugfs

tune2fs, or dumpe2fs (not debugfs) for the stats.

tune2fs -l /live/boot-dev/antiX/rootfs
dumpe2fs /live/boot-dev/antiX/rootfs

[Member marcelocripe]

Helpful

Up

0

::

Hello everyone,

Sorry for the delay, yesterday I managed to start the tests.
I will attach in this post the file “Tentativa_de_Manutencao_da_Particao_no_antiX_14-04-2021.txt”, it includes the texts translated by the internet translator into the English language above my original texts in pt-BR, including the terminal texts that are in pt-BR.

Robin, I hope I have managed to apply all your explanations correctly.
Thank you very much for all the teachings and the tutorial.
I am also grateful for your and Skidoo’s explanations of the correct way to ask the question about “how to recover the damaged file system in antiX”, the wrong way to express myself are the requirements of the other operating system, many years were trapped in the windows (Long live freedom!).

marcelocripe
(Original text in Brazilian Portuguese)

———-

Olá a todos,

Desculpem-me pela demora, ontem eu consegui iniciar os testes.
Eu vou anexar nesta postagem o arquivo “Tentativa_de_Manutencao_da_Particao_no_antiX_14-04-2021.txt”, inclui os textos traduzidos pelo tradutor da internet para o idioma inglês acima dos meus textos originais em pt-BR, inclusive dos textos do terminal que estão em pt-BR.

Robin, eu espero ter conseguido aplicar corretamente todas as suas explicações.
Muito obrigado por todos os ensinamentos e pelo tutorial.
Eu agradeço também pelas explicações suas e do Skidoo sobre a forma correta de formular a pergunta sobre “como recuperar o sistema de arquivos danificados no antiX”, a forma errada de me expressar são os requícios do outro sistema operacional, foram muitos anos aprisionados no janelas (Viva a liberdade!).

marcelocripe
(Texto original em Português do Brasil)

Attachments:

1. Tentativa_de_Manutencao_da_Particao_no_antiX_14-04-2021.txt

[Author]

Posts