Recover SNAP Server 2200

[NetExtra]

I had a client delete files on their 2200 and ask for help. We took it into the shop and before I knew it, my tech had pulled a drive, hung it in a W2K bench machine, and run diagnostics on it. I said I didn't think we could do that and had convinced the client that we needed to just leave it alone. I brought the 2200 up on the single drive and all had looked good. I left for other issues and said to put it all back together. When we put the 2nd drive back in the 2200 it would not come up. I found out that after I had it running before this, my tech ahd tried recovery on the other drive also. I said that should not matter if he did not write to the drive. He said he had no, but then said that Windows did write a signature to the drive. Is this the problem, and is it repairable? I tried and explain with as much detail as possible. Thank you for your help.

[bitor]

See "Snap-tech" on this forum for data recovery. He is very reasonable.
Snap-tech http://forums.procooling.com/vbb/member.php?u=1429
Send him a message concerning your problem.
If for some reason you cannot get a hold of him the last person he helped was admin69.
http://forums.procooling.com/vbb/member.php?u=8335
Write him and ask him how to get a hold of snap-tech.

I believe once anything in windows writes to the snap hard drive you are in deep trouble. Best thing to do is try the recovery method via snap-tech if you have important information on the drive.

Best of luck,
bitor

[NetExtra]

We backed up the data before trying any of this so it sounds like our only real option is to reinitialize the 2200. Is that true? If so, what is the easiest way to achieve that (probably need to install SnapAssist)?

[blue68f100]

If you can boot from one HD you have several options. The best option since drive 1 HD is messed up (from a Windows PC) is to break the array down to a single HD. The FAQ's have these procedures. You need to look at the System log and see if there was a HD failure. SnapOS is very sensitive to timming, so if the SMART had to do multiple/consective repairs can cause this. The reason I test all HD with SpinRite.

The only utility that I know that can do repair on a SnapOS HD is SpinRite by GRC. This program does not look at the data, it looks at sectors/tracks at the bit level, so it does not inturpit the data.

[NetExtra]

What if I just reset the 2200?

[NetExtra]

I repositioned the hard drives and got it to boot, but I cn't rebuild the mirror. How do you determine the drive ID (the 10000 and 10010 as shown in the FAQ. I do have errors on the functioning mirror drive that show it as drive 50000 but I do not know what the other drive is. Is there another way to rebuild the mirror?

[blue68f100]

Yes, there is a way to rebuild the array. But if you get the syntax wrong all is lost. The reason I recommend booting up from a single drive and confirm that ID is 10000. If so it is the first boot device. When you break the array to single HD's NO data is lost, provided the good drive is Master (10000), co de config individual dev 50000. Now you can remove the drive in question and replace. Then from the gui you can build the raid1 array again.

The reset to factory will do nothing for you but make the process more complicated. Does not hurt but does nothing in helping you clear your problem. The factory reset is usefull if you are getting panic alarms.

I think the "info devices" cmd will give you drive info. it's been quite a whils since I've messed with the SnapOS hardware.

[zorka]

Hi my snap server 2200 had been working fone for sevferal years until one day I received this email error as administrator.

================================================== ==================
01/12/2010 16:40:07 ERROR File System Check : Partially allocated inode I=5440
================================================== ==================

The previous message occurred in the following context within the system log:

01/12/2010 16:40:07 INFORMATION System Initialization : Initialization Complete! Memory to be released: 47988816 bytes.
01/12/2010 16:40:07 ERROR File System Check : FSCK fatal error = 27
01/12/2010 16:40:07 ERROR File System Check : Partially allocated inode I=5440
01/12/2010 16:40:06 INFORMATION File System Check : ** Phase 1 - Check blocks and sizes
01/12/2010 16:40:05 WARNING File System Check : partition is NOT clean.
01/12/2010 16:40:05 INFORMATION File System Check : Executing fsck /dev/rmirr0 /force
01/12/2010 16:40:05 INFORMATION File System : Opened FDB for device 0x10016
01/12/2010 16:40:04 INFORMATION File System Check : partition is clean.
01/12/2010 16:40:04 INFORMATION File System Check : Executing fsck /dev/ride2g /fix /fixfatal
01/12/2010 16:40:04 INFORMATION File System : Opened FDB for device 0x10006
01/12/2010 16:40:04 INFORMATION File System Check : partition is clean.
01/12/2010 16:40:04 INFORMATION File System Check : Executing fsck /dev/ride0g /fix /fixfatal
01/12/2010 16:39:58 INFORMATION INIT: Setting IP address to 192.168.0.7
01/12/2010 16:39:58 System Initialization : Server v3.4.790
Build Date: Mar 21 2002 20:16:25
Boot Count: 93

Next thing I know the server shuts down by itself. When I restart my netwrok shares are not ehre but all my users rights etc are listed. The Disk utilitily says there are no disk.

How can I recover my data I had it as Raid1 configuration. I ordered an identil HD.

Any assistance would be appreciated.

Thx

[blue68f100]

If it's not seeing any HD's it sounds like a hardware failure. The ID's given are for both drives, one indicates may be bad.

Try unplugging the power from one drive and see if it will come up. If not try the other. May have to swap position with the HD. Just mark the HD's so you know there original position. With raid1 your data should be good on 1 HD and a min. If it's a MB related failure you will need another 2200 to access the data. Last results would be a recovery service but I do not think you are at point yet.

[zorka]

Thx blue68f100

Would a snap 1100 be able to see the data? I have one already available? How do you get one snap to see data on another?

Thx for your help


Will try your suggestions. Have oredered identical HD incase one had a failure.

[blue68f100]

If it's a newer 1100 v2 it should work. Snap use to use the same MB in both 1100 and 2200, just did not populate the board.

What you need to know is make sure the 1100 v2 is one that does not have OS in ram (v1).. Other wise it will install the os to match it's hardware. Normally it does not wipe the data but there is always that chance.

[zorka]

Hello blue68f100,

Just returned from a trip and checked the 1100 data.

Its as follows:

Model Software Hardware Server # BIOS
1000 series 4.0.855 (US) 2.0.2 201222 4.0.855

Would this 1100 be able to read a 2200 snap server?


Thx for your valuable help.

[blue68f100]

By the platform byte it indicates to be a v2.0.2 1100. A good test is to see if you can access the admin pages without a HD. If you can not it is a 1100 that can read 2200 drives. Now the 1100 has a new OS, which should not come into play since it boots from the HD.

[zorka]

Thx blue68f100

Just to make sure I understand. Test would be to remove drives from 1100 and try to see if I get to see the admin pages? Pls confirm.

Supposing theis 1100 can read the 2200 drives how exactly does one get the 1100 to read the drive? Install the good disk in to the 1100?

Thx again for your help