What is my 4100 doing!

[Phoenix32]

Originally Posted by bug1124

I don't suppose anyone has a Quantum LM30A011-01-A here locally that they would like to sell me...

I have a couple of those, but I am not local and to be honest I do not think that would be your best bet. If one drive is failing, the others are probably not too far behind. Get NEW drives...

[bug1124]

I followed the directions regarding the network shares and got the following:

The path "private" does not exist.
The following subdirectories will be created:
"private"

Do you want to create this path?

So you think I should skip buying the replacement quantum drive and go ahead with four 120GB? I read earlier that I need to match within 2% the capacity...do they even make 30GB drives anymore? What is 'live DD'?

[blue68f100]

I would skip the path. I'm not sure that is the correct way to go.

Well bug, dd is digital dolly. It is a unix/linux utility for reading and copying hard drives. In order for my suggestion to work you will need to install the new drives and build a empty raid 5 array. Then you must read the mbr and the partition table. And compare it to the original drives. Then copy the partitions to the new dirves. This must be done for every drive. And there is alway a chance it will not work.

If you have a copy of spinrite, it may beable to repair the drives. It can atleast read the smart data.

[bug1124]

Well I have DD and SpinRite. I have made a SpinRite boot disk and burned DD to a CD. You said imaging the drives will be our last alternative, so what is the first step? Which is my best alternative and what should I make sure NOT to do.

I am going to convince the owner to let me take the Snap unit home for the weekend and attempt to repair it. If you think buying 4x 120GB drives is the way to go I can also try and get that cleared.

On a completely unrelated note...if anyone wants to play with a sick 4100 this weekend there will be free beer. PM me for directions!

[blue68f100]

Sorry, my medication doesn't allow alochol.

My first step would be to make a image backup of the drives. Normally this is done to identical drives. With different size destination I would just make a copy of all the partitions. 1 folder for each drives partitions.

I would get the 120's and have the 4100 build a raid5. You will need them since you have filled the 30gigs up. Then read the MBR and Partition data. With that info you can compare to the original drives. Actually you can copy the partitions back with out affecting the MBR. You have to use a special cmd for that. Make sure you mark the positon and/or number if the drive clearly. If you cross 2 drives, it kills the array. It's possiable we may only need the data section 60000. Never have done this before it will be a good exercise. I havn't started failing my 4500 to see what the recovery option will be.

[bug1124]

So if I understand this correctly I need to do the following:

1. Remove the existing drives from the Snap unit carefully marking their drive number on the top.
2. Install 4 new drives (4x 120GB) into the Snap unit and format the array as if I was starting fresh.
3. In another computer install one hard disk from the Snap unit along side a non bootable drive (formatted fat32 that has more than 130GB free).
4. Using DD (or is it SpinRite?) create an image (.bin?) from the Quantum drive on the 130+GB drive.
5. Using DD (or is it SpinRite?) create an image of the MBR (file type?) from the Quantum drive on the 130+GB drive.
6. Repeat steps 4 and 5 for each of the four Quantum drives.

After this point I am a little sketchy on what we are going to do, but if the above is correct I can get started. Two questions though...1) Am I using DD to create the images on the 130+GB drive? 2) What format am I supposed to back up the MBR in?

[blue68f100]

Step 3. Do not copy to the new 120 gig drives. You need to save it to another location. When you use dd to copy the partitions. Use the NAME that is on the original drive.

DD supports Network Shares if you need a place to send the files to.

When you copy the partitions do not use the ".bin" it will assign it's own.

You will need to send me the MBR of both new and old dirves. To do that drop to cmd line. The cmd will look like this "if=/mnt of=/mnt2/org.mbr count=446 bs=1" Where mnt is the source and mnt2 is destination. I need to verify the only difference is the data part.

We will talk after I get the MBR to look at.

[rpmurray]

The listing for the in dev command is missing Logical Device 1000E which should be showing between the JBOD at the top (Logical Device 10006, Position 0) and the RAID below (Logical Device 60000, Position 1). As far as I can tell from my unit, it looks like that would store a duplicate copy of the system files on Drive 2 (drive slot 1). Anybody have any ideas on why that would be missing? Looking at the listing it doesn't look like Drive 2 is dead, so I'm wondering what happened here.

The thing that really worries me is that it doesn't seem to be getting past Phase 2 on the last few times it's tried doing the disk check. That seems to be because of some damaged inodes.

My guess is that this unit was set to run the disk check "only when there are suspected problems" when starting up. And until he ran the disk check, nobody had tried to access a file that had a problem, with PANICs apparently not considered a problem that would trigger a disk check (I can't tell now because the original log file seems to have been replaced with the updated one). But now that the disk check has been run manually, it's discovered some problems and won't continue on to mount the raid.

In my case I ran the disk check option to repair the severe errors and that at least let me back up most of the files, while only losing some non-critical (to me) ones. I don't know if I'd recommend that path here until he's tried everything else, because that might lose him the very files he's trying to recover. If those files are VERY important then it might be safer to send it to some recovery experts. They have a lot more sophisticated ways of recovering the data, although it will probably cost an arm and a leg.

The above is just my opinion.

[blue68f100]

Does anyone know the cmd to turn off file system check on boot. I seam to be over looking it.

FYI, The live cd of dd did not reconise any partitions on the snap drive. I was expecting to see the drive partitions but not any thing else. The Guardian OS file structure is view able.

[snap-tech]

As I have said many times before!!!!


There is nothing that is going to see any kind of file structure on a snap os other than a snap.



The quardian os is based on linux/xfs and can be seen by certain utilities.


I will be setting up a new domain deadicated to the Snap product line within the next couple of months. All information provided will be fact and not quess work.

Until then, be carefull with the suggestion that are being givin out to people who have issues with a snap. You can do more harm than you realize.

Douglas

[rpmurray]

Originally Posted by blue68f100

Does anyone know the cmd to turn off file system check on boot. I seam to be over looking it.

I don't think there's one that will completely turn it off. There are two settings in Disk Utilities > Disk Startup Settings:

Perform a thorough disk check operation:
Always
Only if there are suspected problems

[re3dyb0y]

Originally Posted by snap-tech

As I have said many times before!!!!


There is nothing that is going to see any kind of file structure on a snap os other than a snap.



The quardian os is based on linux/xfs and can be seen by certain utilities.


I will be setting up a new domain deadicated to the Snap product line within the next couple of months. All information provided will be fact and not quess work.

Until then, be carefull with the suggestion that are being givin out to people who have issues with a snap. You can do more harm than you realize.

Douglas

That sounds useful

Please keep us informed

[blue68f100]

The new FreeBSD v6.1 can read the XFS file system now. I'm going to set my 2200 up as a std share (No raid) when I get the 4500 in operation and verify it. But it will not help with a raid5 array. But it may be away to read the 1x00 model.

I talked with bug last night and the unit finally showed drive 3 had failed. Which is not the 10004 we were seeing in the log. Just got a fealing that 1 has failed, another may also be, and the disk may not have enough free space for cache. We terminated all operations/boots with the original 30 gig drives, soon as the snap indicated a failed drive. And concluded that it needs to go to a recovery service.

Snap-tech, you offer a lot of negative critism. But you fail to answer any of the PM I sent you early on. Just because you call your self a snap tech, does not prove any creditability. You have yet to contribute to this forum. All you have been trying to do is get work. As far as I now no one from the forum has used your service, or have not posted any feed back.

We have know early on that the Snap OS uses a modified XFS file system. But we all know that a Raid 5 is suppose to run in degraded mode with only 3 of the 4 drives. We were trying to get degraded mode working to recover the data. If you know the cmd to stop the file system check, we would like to know what it is. And we would like to know how to force up a share. The log isolated were the main problem directory/share is. Would just like to mount several shares if that is possiable, since the 60000 (raid5 array) would not mount. Was wanting to know if there is a work around.

Don't get me wrong this forum is used to transfer info between users and help solve problems. If you have something to contribute, we welcome you.

After talking to him, he know what he's talking about. And I'm glad he is on this forum. He has extensive expertise in HD & HD Arrays recovery for Snaps. He use to man the phone for tech support as a lead engineer.

[snap-tech]

Blue68f100,

ok you want proof that I was the lead technical support person for snap?

lets see, how do you want that proof. Would you like me to paste a copy of my w2 up. would you like me to a directory listing of the source code that used used to create the snap system? for now I will take a picture of my business card from snap and paste it within next 5 minutes.

standby

[snap-tech]

also,

you have no idea how many people i have helped through forum. mY personal email address is not hidden. I have helped dozens from this forum over the years.

I will have put there emails together and forward them to you.

I have made suggestion when needed and correct. And have made negative statements when ever they were as well needed.

at this time I am putting a little juice into camera battery, but will have the picture up in a few minutes.

Only after I know that there only hope for getting there data back is data recovery have I made the attempt. And then at approx more than 75%-85% of the cost that any data recovery company would charge.

more to follow -- I need a smoke break damit all to hell

I feel you have done me great injustice. You have no idea who I am or how many of this forums users I have helped over the years. You have to realize, I cannot supply all the information / utilities that I have. Snap would have my A??. and It would hurt me down the road.

ok now I am going to have that smoke -- be back in a few minutes

[snap-tech]

To blue68f100,

no I have finished my smoke. I came in to ask you to call me toll free at 1-866-279-2985 as soon as you see this. I think we need to get a few things straight and is much easier than trying to type it all here.

so please give me a call. It will benefit you as you can ask me specific questions I will be able to answer them.

waiting for you call

Douglas

[blue68f100]

I just ask that you provide user input to the question we are unable to solve, will be proof enough. Paper means nothing to me. I have worked with PHD's all my life, all it means is they are capable of learning. Most when out side there field of expertise could not give you the time of day. Then there were those few that were execptional. I worked with the best of the best in a R & D lab. Most of my projects came to market 5-10 yrs.

Do you have a patch to allow the 4100's to support LBA48bit????

Please answer this for me ... It should be something your fammilar with.

1. RAID 5, 4 x 30 GB drives (all the exact same)...
2. Store Test Data on the array.
3. Pull the power plug on drive 2.
4. Refresh Disk Status display (web interface menu).
5. Nothing.
6. Check error logs.
7. Nothing.
8. Go to Server Setting (web interface) and tell it to do a system
restart (forces a disk check).
9. SNAP comes back online, go check Disk Status display.
10. Shows Disk 1 as orphaned (the unit was not even shut down, just a
restart) and the RAID as broken.
11. Shutdown unit (it does a proper shutdown).
12. Plug Drive 2 back in, power up the SNAP.
13. Unit comes up, go to Disk Status display
14. Drive 1 Orphaned, Drive 2 Orphaned, RAID Broken.
15. Logs show 2 drives missing from RADI set (duh).
16. Format Drive 2 (simulated bad drive).
17. Make Drive 2 a Spare.
18. Nothing.
19. Reboot the SNAP.
20. Nothing.
21. Drive Staus display still shows Drive 1 Orphaned, Drive 2 as a
spare, and RAID Broken.

[blue68f100]

I feel you have done me great injustice. You have no idea who I am or how many of this forums users I have helped over the years. You have to realize, I cannot supply all the information / utilities that I have. Snap would have my A??. and It would hurt me down the road.

Snap-Tech I did not mean to stirr your blood pressure.

After talking with him for 1hr plus, he is a very very very qualified Snap Tech. His knowledge of all Snaps far exceds mine. He has over 12 units. ( Hallis you have a way to got to catch him.) He is just so busy doing recovery work and starting up his bussiness, he normally does a direct contact. As we all know phone support is faster than this forum. He's the real McCoy.

His expertise is in HD and HD Arrays Recovery.

Phoenix, The problem posted is actaully a bug in the 4000. But the fix is easy. Change the drives to C/S like the later models (rev 3+) and see if that fixes it. He also said the data is not lost. If you were to do a raw copy to new drives and install them, It would reconize the array and start the resync. Apparently the snap has a record of the drive ID's.

[Phoenix32]

Now if I can get my hands on some C/S cables without paying like $10 each for them...

Leave it to me, I WILL find the variouis bugs where possible...


While we got him here, there is one question I think is BEGGING for an answer...

Snap-Tech

-OTHER THAN MS AD 2003-

What changes/fixes are in SNAP OS v3.4.807 and SNAP OS v4.0.860 from SNAP OS v3.4.805? Surely Adaptec can't get pissed for answering that one I would hope. This one has been a puzzle for some time now. Can you clue us in please with more than just a "various bug fixes" answer that we always get?

And oh yeah, as a side note, can the 4000 support more than 256 MB SDRAM?

And while I am at it, where at in Washington? I too am in Washington. Just curious...

[blue68f100]

Update on v2 failing Orphaning drive 1 on drive 2 failure. Phoenix made CS cables and changed jumpers to match. This did fix the bug in version2 4000.

Good call Snap-Tech.

[Phoenix32]

Yes, thank you Snap-Tech, you saved me probably 8 hours or so of headaches. I would have figured it out sooner or later (I have a v4 4000 also and was going to test with it next), but the extra testing and so forth and figuring that part out would have taken hours. THANK YOU! I meant it...


Dave, I would tell these guys what the problem is, what the fix is/was, and why, but I am a butt head and they wouldn't break their 4 x 320 GB RAIDs for us to find out if they would rebuild properly, so I am going to let them all find out the HARD WAY. Then when their data is gone, I will tell them all about it...

Aint I rotten...

[bug1124]

Well I wanted to update everyone on my situation. I have spoken with Snap-Tech and decided to have him recover my data as well as upgrade my unit with new memory and hard drives that I am shipping to him.

I will post the results of the recovery/upgrade. He GUARANTEES 100% recovery from my system so I am very hopeful.

BTW thanks everyone for the help on this issue!

[blue68f100]

Thats good news for the recovery.

We try to help any one in need.

[jontz]

Man....I've been gone too long...I missed the great blue vs snap-tech face off!!!

[snap-tech]

To let everyone know. Bug1124 sent me his snap and I was able to do a full 100% data recovery at an excellent price.

He has his snap back and all is good. He should be posting some soon to support my statement.

Douglas snap-tech