Topic: corrupt files

[jameswilson]

i have a poorly sme 8 server with a degraded raid 6 array. I have hardware built ready to replace and i was using affa to back it up nightly up until 3 days ago. Today about lunchtime users started reporting issues with files. Affa isnt working as it fails, and i cant copy files using a pc to copy the ibays. However ftp seems to work. I dont want to restart it in case it doesnt come back. Im guessing i can restart some services like smb to get these to transfer again and replace it tomorrow.

Any ideas please

James

[guest22]

Im guessing i can restart some services like smb to get these to transfer again and replace it tomorrow.

You didn't stop services manually did you? So why start smb, for there is another reason smb is not functioning.
I suggest you be very careful with what you have left right now and make an image of the disk as it is now, and shut it down now.

One thing is for sure, you won't be able to satisfy your co-workers today.

The wiki holds info on how to further address this issue, and I hope either stephdl, terryf, janet or stefano can help out. But you need to make a disk image and shut t down, or at least log off all users. Did I say NOW! ?


guest

[jameswilson]

i cant image a 5 disk raid 6 array with 2 fails or can i?
i have got most of the important stuff over ftp, so the fact ftp works but smb doesnt makes me think the data and the degraded array are still ok. ie its something else.
if ftp transfers work 100% and everything is apart from networks access on the windows machines and affa im hoping i can get an affa backup off it before i try a reboot?

[guest22]

I don't know if "ftp works 100%" means the data that it transfers is 100%

[jameswilson]

yes agreed but wont transfer via smb

[guest22]

You need advice from someone more skilled then me in this matter.


Shut it down, will it boot? I don;t know. What IF it does not boot correctly? What can I save now and what way? Can I make an image.


Anyway, AFFA does not need SMB

[Stefano]

jameswilson, you are saying "affa is not working" but we have no evidence.. no log, no errors, nothing.. we can't help you without knowing what's going on..

[ReetP]

I believe you can suffer two failed drives on a RAID 6, though the rebuild time may well be long. Why didn't you put in a spare when the first one went ?

First as pointed out you need to supply some more information on your errors.

To see what is happening on your raid for starters :

cat /proc/mdstat

affa - status
affa --list-archives

AFAIAA Affa uses ssh/ rsync so shouldn't be affected by SMB issues ?

What errors have you got in your logs ?

/var/log/messages
/var/log/affa.log

Have you got spare disks sitting there ready to go ?

Can you ssh / scp a copy of your data to another machine as an emergency backup - the wiki has some notes on the files you need (or check your affa backups for a list of directories files). Don't forget a copy of your mysql databases.

I think the most important thing is to get stop your users doing any more, and get a good backup if possible. They may thank you in the short term but.......

You can then sort out trying to resurrect your RAID, failing which you can rebuild from your backup.

B. Rgds
John