NASLite Network Attached Storage

OK so i had a brown out which caused no end of grief to my kit a few weeks ago.

All was fine but when i went to reboot my NAS box one disk failed to come up due to errors.

I have used the scan and repair tool twice and whist it seems to try fixing things it returns Warning ***Filesystem Still Has Errors***.


I am about to remove the disk and mount it under Knoppix or the like and try from there but can anyone recommend a more formalised procedure for doing this so i dont compound the error?

Worst case scenario one of my clients has R-Studio i could use but its a 500GB SATAII drive and will take an eternity to recover using this approach.

Thanks for any tips at all.

An update. Dropped the disk into a Kanotix box and ran the same commands as Naslite did and the disk repaired itself.


OK so i now have loads of stuff in lost and found however can someone explain why the same tools fix it on another OS.?

I've been running Linux since 1993 on a personal level, and in a production enviroment since 1996, so I have literally seen fsck do it's magic more than a few times. Generally fsck will repair on it's first run, however I have seen several times it's took 1 to 2 reboots and rechecks before the file system is clean, and in some instances (after many reboots) I had to boot a slackware cd to fsck the server.

Let's keep in mind that the live filesystem did reside on these drives and I would imagine a critical area of the disk had been compromised for it to take so many fsck's. In your situation I would say what your describing is fairly normal. Where you reading/writing to the disk when the power problem occured?

i cant be sure about reading/writing but its a safe bet that i was when the power started popping.


I know what your saying about fsck and i cant quantify this statement but my gut is that i seen the same inodes listed on both pass 1 and 2 or NASlite and pass 1 of Kanotix.

The difference being after pass 1 on Kanotix they went away.


I wish i had taken lots of notes but i didnt expect to have to do this so i wasnt prepared.

Perhaps it was a glitch in the matrix but my unquantified gut is that Kanotix worked better than NASLite using the same tools.

have you considered adding a UPS for your server to prevent this from happening again?

i already have 2 but i have ordered 2 more to cover the rest of the kit

I get this error as well, file system listed as 'not clean'. I went to to the administration utility - storage configuration and the 'check and repair' option has an asterisk, I cannot select it. I can do a check, but not a check and clean.

Any idea why I cant repair? Do I need to remove all data and reformat to clean it?

Thanks much

Try option 7 Apply Journal. that should do the trick and will not damage data or need to format.

I installed a disk from my old NasLite+ 1.5 into my new NasLite V2.04.
The IDE port died on the NasLite+ 1.5, so that is why I moved it.
I installed it and applied Journal on it.
Now looking at the disk hardware details I see
Filesystem features: has_journal filetype sparse_super large_file
Filesystem state: not clean
Errors behavior: Continue
I ran the "Check filesystem" but no errors were reported.
I can't select Check and repair Disk-0 filesystem as there is a asterisk on it.
Is there another way of fixing or cleaning this disk?

This is not an error. When the filesystem is mounted it is flagged as "not clean". That flag will be reset to clean when the filesystem is unmounted at shutdown. Next time NASLite reboots, if the filesystem is flagged as clean then no check will be forced.

Of the machine has problems and goes down due to power failure or alike, then the filesystem will not be properly unmounted and therefore will not be flagged as clean. In that event, on reboot a check will be forced since the filesystem will be flagged as "not clean".

In other words, when the filesystem is active, it's always "not clean".

Hope that makes sense...

Yep that makes sense.
Thx

After a reboot I now get
Filesystem features: has_journal filetype needs_recovery sparse_super large_file
Filesystem state: clean
Errors behavior: Continue

What does this mean?
"has_journal filetype needs_recovery"

Thx again.

The description I posted above was actually describing the "needs_recovery" flag. I apologize for the confusion.

The not_clean usually means that there is a discrepancy. In most cases the filesystem will be flagged as not_clean if you delete the lost+found folder. Generally a filesystem error is resolved at boot time.

I have a drive that has suffered from being in a NASLite 1.0 box, and there was a power problem.

So I read all of the related posts, that I can find.

I have built a Kanotix disk, and wanted to know what the command was that the user called 'fat' used to recover a disk in a similar state, to mine.

I thought that it was the fsck command that was used, and as I have no idea about the structure of these commands, any help would be greatly received, or are there any other methods / tools that I should consider?

The other drives have all been transplanted into a NASLite 2.0 box, and they have been recovered, but this one stops NASLite 2 coming online.

This is a great product, and next step is an online UPS.

Thanks in advance

Artful

Let NASLite-2 scan the drive and attempt to fix it. It may take some time but it does work. If it does not, then you can use the manual filesystem repair option to do a complete check and repair including bad blocks. All you need is your NASLite-2 to get the drive back up and in order. If you have enough RAM in the server, then you should have no problems getting back in business.

That's how I'll go at it anyhow.

dimension

Thanks for that I will give it a try, it just seems to take an extra long time to mount this drive. But if you think that it will eventually mount I will try it again. One of the drives did mount and I was able to run the utility manually and it did recover. However I eventually removed this drive from the box, and I had left it on for more than 24 hours, and it still never finished. There is a couple of hundred gigs of stuff on the drive.

I wasn't expecting it to take that long, the other drives all recovered after about 20mins each.

Cheers

Artful