horrible problem with .tgz backups

[WoAyumi]

Dear NFO,

I have a horrible problem with all my backups. All archives are broken. I am receiving all kinds of errors on different archives:

I need help to restore at least one archive. Help!

error1:
tar: Skipping to next header
tar: Exiting with failure status due to previous errors

error2:
gzip: stdin: invalid compressed data--format violated
tar: Unexpected EOF in archive
tar: Unexpected EOF in archive
tar: Error is not recoverable: exiting now

[Hot Carl]

What commands are you using to tar?

Carl

[WoAyumi]

Code: Select all

#!/bin/sh
FILENAME=bak-World2-$(date +%Y%m%d-%H%M).tgz
cd /mc/worlds
nice -n 5 tar czvfsp $FILENAME /mc/worlds/World2
mv /home/mc/worlds/$FILENAME /spinner/bak/World2/$FILENAME
exit $?

[Hot Carl]

Pull the z out or use tar -zcvf, or just use tar -cvf

See if that works.

Carl

[Hot Carl]

Pull the z out or use tar -zcvf, or just use tar -cvf

See if that works.

Carl

Oh wait not sure if it matters or not, but there might be an issue with the .tgz extension. I think that you have to use the "z" with the .tgz extension. Is there any reason why it can't be .tar?

Carl

[Hot Carl]

I can't edit my posts, sorry for the 3 posts in a row but you need to use tar -zcvf that will compress to extract it you would use tar -zxvf.

Carl

[WoAyumi]

I'm using zxvf to extract. It does not help. What should I do with those backups? They all showing errors.
Why is this happening? What caused these problems?

[Edge100x]

Are you certain that there weren't errors when the backups were created?

[WoAyumi]

How can I be? It's an automatic script, that initiates by itself.

How can I restore at least one archive? Help. How to not make sure, this will not happen again?

[Edge100x]

Automatic scripts normally have their output to an email (when run via a crontab) or log. Did you also not test the script before you started automating it, though? That is not good.

I do not know ways of recovering from the errors that you are seeing.

[WoAyumi]

Edge100x, where are these logs, normally, in debian?
Yes, I am using crontab.

[Edge100x]

Run via crontab, you should normally see output emailed to your local root@machine account.

[WoAyumi]

Where exactly this file can be found, by default? I'm sorry.

[WoAyumi]

this is, what I was able to find:

Code: Select all

Jul 28 12:00:01 ispconfig /USR/SBIN/CRON[31964]: (root) CMD (/mc/bak-World2.sh #World2 bak)
Jul 28 12:06:15 ispconfig /USR/SBIN/CRON[31958]: (CRON) error (grandchild #31962 failed with exit status 1)

Can this help to find out, what was the cause of corrupted archives? Is this information helpful?

[Edge100x]

That suggests there was an error, but doesn't indicate what it was. You could likely replicate it by running your backup script manually. But, it will still likely be difficult or impossible to recover an old and broken backup, even with that information =\.