Snap 4100 orphaned drive question

[blue68f100]

I remember seeing that on my 2000, I think it is in reference to the panic alarm.
If you are able to look at the log, Means the 4100 has booted, but may have not mounted the array or with errors.

You are right with the error. It's pointing to a memory/cpu register.

Did you reseat the ram?
Startup the unit with the top off. Make sure the cpu fan is working if equipped.

It may not like the 80 gig drive at all? There have been many instances where snaps did not like drives. And the raids are pickier.

[Reason_Man]

I figured we knew the server booted properly since I posted all those things from the individual drive logs in the first post.

I have reseated the RAM, i've unplugged every cable from the board and drive and then back in. The only thing I haven't pulled is the CPU and the battery. All the original drives are back in and plugged in as of the post with the panic error. The top has been off of the server since I started work on it three days ago.

The first thing I noticed when I cracked open the case was the CPU fan doesn't constantly spin but spins at seemingly random times. I just chalked this up to it only spinning when it needed to like other fans i've seen that only spin up when it gets to a certain temp. If that is not the case in this situation, please let me know.

[jontz]

You are correct about the CPU fan; it only spins up when the CPU gets hot enough that it needs to. This is perfectly normal.

The more you describe your problem, the more I am thinking that you might just have some bad RAM, especially after you posted the general protection fault error. I would try swapping the RAM out if you haven't done so already. Either that or head over to Plano and show up on Blue's doorstep with your snap in your hands and a sad look on your face.

I know that it would be nice to have another working 4100 to test some of these theorys out and/or rebuild your array, but sadly eBay is the best place to source a used 4100. I got mine for $112 with no drives in it, so it may not be as bad as you think price-wise. I wish that I had some more/better ideas for you...if I think of anything I will let you know.

[jontz]

Originally Posted by Phoenix32

ok, which is why I hinted I was not sure on a 4100. Is this true for all 4100s? Asking because some/most 4000 use CS (like mine), but some do not.

Yep, all the 4100's are the same...they have 4 controllers instead of 2, so all drives are set as master.

[blue68f100]

Yes the fan is temp controlled.

I think you need to find a replacement 60gig drive. Does the drive status show more than 1 drive missing?

[Reason_Man]

Well, I replaced the RAM and the log is showing ALOT more activity than it has in the past, like disk errors and fixing and shit. I'll post more when it gets to a stopping point.

EDIT:

A bunch of entries for things like :

File System Check : -2070463649 Bad I=16613630

and

File System Check : Unref File I=10415351 Owner= Mode=67775
/dev/rraid0: Size=1970893072446623458 (Reconnected)

Now it just moved into phases 2 3 and 4 of the disk check, further than it's gotten before.

EDIT:

Well, it just went back into the part where it looks like it's frozen. I did a scan of the devices on my network before and after the "freeze". Before the scan shows the server. After it's gone.

[blue68f100]

This is reporting raid 0.

/dev/rraid0: Size=1970893072446623458

Raid 0 (stripped) = NO REDUNDENCY ALL DATA LOST

jontz, notice the size of raid0, 1.9 x 10 to 18th. This may be the magic number that has kept us at bay.

[Reason_Man]

Ok, so this means i'm boned right?

Could you explain this "magic number" and its significance?

[blue68f100]

Don't no, if you need the data your only chance of recovery is through a service. They have parts for all drives and have special hardware for reading the drives.

1 other way to confirm this is you see reference to disk id 40000 = raid0, 50000= raid 1, and 60000= raid5 in the log.

We have been trying to figure why we could go larger than 1-1.2T on the 4000 units in raid5.

[Reason_Man]

Everything i've seen has always been 60000.

[Phoenix32]

Originally Posted by blue68f100

For your information on 4000:
SN < 70700042-003 use M/S (40 ribbon connector)
Version 3 and > CS (80 ribbon connector)

Yup! And mine is a -004

[Phoenix32]

Originally Posted by jontz

The more you describe your problem, the more I am thinking that you might just have some bad RAM, especially after you posted the general protection fault error. I would try swapping the RAM out if you haven't done so already.

Took the words right outa my mouth...

[Phoenix32]

Originally Posted by jontz

Yep, all the 4100's are the same...they have 4 controllers instead of 2, so all drives are set as master.

ROFLMAO I knew that too...

[Phoenix32]

Originally Posted by blue68f100

jontz, notice the size of raid0, 1.9 x 10 to 18th. This may be the magic number that has kept us at bay.

jontz only! I feel left out... What about me? LOL


Ya know, you might be right. Now what did I do with that math book?

[blue68f100]

Originally Posted by Reason_Man

Everything i've seen has always been 60000.

When you no a "co de info" does it referrence to 60000. and only shows 1 drive missing. I'm thinking you may have failed a ide port, with the panic alarms.

If every thing you have seen is 60000 you proably is in raid 5. The error message was strange in reporting raid0 unless it was refering to drive 0 of the raid set, which happens to be the boot drive. If it going into panic mode during the scan it a hardware problem. You basicly have 3 options. Find a replacememt drive and see if it like it. Buy another 4100. Or contact one of the recovery companies and send your drives in.

[Reason_Man]

I put in the spare drive with the new RAM and it seemed to do more this time than it has before but ultimately "froze" again. I refreshed the log page until the last minute of life and saved it in case you could pull anything from it. The dates say 1999 because I inadvertently reset all the system settings instead of just the admin password. The log starts yesterday and picks back up at about 830 today. Looking at it now, checking doesn't seem to get past phase 4.

Click

EDIT:

Looking around on the wiki, I found this. If I put back the original orphaned drive, would that work or no?

[blue68f100]

Looking over the log it looks like drive 2 (10008) was formated. All of the problems accoured when it did a check of raid 5. Had to many errors to repair. The errors just run for ever. Which is more indicative of a hardware problem.

My advice to you would either find another 4100, and/or a 60 gig drive matching what has failed. With a matching drive it may work. I have seen that mixed capacity do not work. The drive gemoetry is all wrong.

Another option Which I would proably do any way, knowing I need the data off of it. I would try to clone the drives to another drive or file. Starting with drive 2 that failed. Does it spin up? Make sure you mark the posion or ID of each drive.

If you can come up with 4 identical drives that are good. I would use Digital Dolly (can be found at Majorgeeks.com or download.com) and clone the drives over to the new ones. I would do a RAW Copy which requires the same size drive. Then install the clone drives back in to the server. If you have a spare 300 gig. You could do a clone to image file. If one of the drive is having a problem it will show up. After you have made a clone of the drives. I would try maybe SpinRite or other disk utility to see if it's capable of repairing the drive. IF YOU ARE PLANING ON USING A RECOVERY SERVICE, YOU DO NOT WANT TO ALTER THE DRIVE IN ANY WAY.

You are some were in texas, as I am. Are we near each other, knowing this state is second only to Alaska in size.

[Reason_Man]

If I can find an identical drive then i'll try again. Failing that i'm just going to have to bite the bullet and either send out the drives or give up(depending on the cost). I've sent out for quotes and so far i'm looking at 5k which is WAY out of reach for our small business. I'd like to do this myself but I just don't trust my ability to do that without further ruining things. If the data wasn't that important than i'd try but considering that it is, i'm not about to chance it unless it's easier than i'm thinking.

I'm in Beaumont, just southeast of Houston on the TX, LA border.

[blue68f100]

I think you have a chance of recovery. The best odds are going to be with another 4100 and a replacement drive. But it may only take the drive. Provided another one hasn't failed in the process. The 4100's have been selling on ebay for around $100 mark with out drives. If you pick another one up you could use a S2S to backup the snaps to each other. This is a common practice. Here is acompany selling the mb for the 4100 http://www.northwesttechnical.com/in...roducts_id=761

If you haven't revomed and reseated the CPU and fan I would do so. I would also do some forced air cooling on the cpu. IF the fan is not running during the heavy activity. I know on my failed 2000, if I keep the cpu and some ram in the area it will boot. But soon as it starts heating up it goes in to panic mode.

[Reason_Man]

I have not reseated the CPU, i'll try that. but it doesn't look like a standard socket so is there a special way to remove it? I don't want to just go tugging on things and make it worse.

EDIT:

Good news! The guy that sold us the server gave me a call back and told me he had an extra 4100 at his house so tonight i'll be taking my server over there and we'll work it out. If I get it fixed and we figure out the problem, i'll let you guys know.

[jontz]

Originally Posted by Phoenix32

Is this true for all 4100s? Asking because some/most 4000 use CS (like mine), but some do not.

Originally Posted by jontz

Yep, all the 4100's are the same...they have 4 controllers instead of 2, so all drives are set as master.

Originally Posted by Phoenix32

ROFLMAO I knew that too...

Then why did you ask, you goober!

[jontz]

Originally Posted by blue68f100

jontz, notice the size of raid0, 1.9 x 10 to 18th. This may be the magic number that has kept us at bay.

Interetsing that it is reporting that number, since he doesn't have near that much drive space in the machine. It does appear to be the magic number though...perhaps this is the limit for the snap OS.

[jontz]

Originally Posted by Phoenix32

jontz only! I feel left out... What about me? LOL

'cause I am way more cooler that you, n' stuff! HA!

[Phoenix32]

Originally Posted by jontz

Then why did you ask, you goober!

Because I FORGOT... LOL DUH, why do you think I used the head bang symbol you goober?

[Phoenix32]

Originally Posted by jontz

'cause I am way more cooler that you, n' stuff! HA!

Prove it...