4100 Goes ballistics

[Fredo]

Hi,


I'm in trouble with our 4100; maybe someone can help me.

At a certain point, the 4100 just stalled. I had to power the unit down to get 'em working (?) again.
At that point, the 4100 reported a bad disc #3.
Having replaced discs before, we swapped the broken disc #3 with a spare one.
Apparently the rebuild went fine,and at first sight all seemed normal.
Until we got "error 33" messages.

The data on the snap was not crucial, so we decided to start from scratch again.
We have formatted the discs, rebuild the Raid, and again ... all seemed fine.
Until we put some data on the disc; pretty much immediately we got a new error 33 or 39.

Then we used 4 discs out of another 4100 which wasn't used anymore; same problem.
Then we bought 4 brand new discs, replaced the battery, performed a system reset, checked connections, checked fans, etc ...
Still the same problem.

===> format discs = OK
===> Build Raid = OK
===> Create folders & shares = OK
===> check disc ===>> error 33 or 39



Attached the latest crash log.
Ideas are welcome.

================================================== ==================
08/11/2008 11:01:07 ERROR File System Check : Bad state 0 for inode I=39005
================================================== ==================

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

08/11/2008 11:01:07 INFORMATION System Initialization : Initialization Complete! Memory to be released: 71920944 bytes.
08/11/2008 11:01:07 ERROR File System Check : FSCK fatal error = 39
08/11/2008 11:01:07 ERROR File System Check : Bad state 0 for inode I=39005
08/11/2008 11:01:07 WARNING File System Check : Summary information bad (Salvaged)
08/11/2008 11:01:07 WARNING File System Check : Blk(s) missing in bit maps (Salvaged)
08/11/2008 11:01:07 WARNING File System Check : Free blk count(s) wrong in superblk (Salvaged)
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 5 - Check cylinder groups
08/11/2008 11:01:07 WARNING File System Check : ACL i-node 525: bad dirlink (256)
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 4b - Check backlinks
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 4 - Check reference counts
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 3 - Check connectivity
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 2 - Check pathnames
08/11/2008 11:01:07 INFORMATION File System Check : ** Phase 1b - Rescan for more duplicate blocks
08/11/2008 10:44:32 INFORMATION DHCP/BOOTP: Setting IP address to 192.168.10.102
08/11/2008 10:44:10 INFORMATION File System Check : ** Phase 1 - Check blocks and sizes
08/11/2008 10:44:08 INFORMATION File System Check : partition is clean.
08/11/2008 10:44:08 INFORMATION File System Check : Executing fsck /dev/rraid0 /force /fix /fixfatal
08/11/2008 10:44:08 INFORMATION File System : Opened FDB for device 0x1000E
08/11/2008 10:44:08 INFORMATION File System Check : partition is clean.
08/11/2008 10:44:08 INFORMATION File System Check : Executing fsck /dev/ride1g /fix /fixfatal
08/11/2008 10:44:08 INFORMATION File System : Opened FDB for device 0x10006
08/11/2008 10:44:08 INFORMATION File System Check : partition is clean.
08/11/2008 10:44:08 INFORMATION File System Check : Executing fsck /dev/ride0g /fix /fixfatal
08/11/2008 10:44:02 System Initialization : Server v3.4.790
Build Date: Mar 21 2002 20:16:10
Boot Count: 5

[blue68f100]

Have you read the notice on the 4100, if so does your 4100 have the mod or newer rev MB. Inodes mean trouble, big trouble. The mod corrects a timing problem with 240gig and larger units. If you have the Snap OS do a re-install. If you have spinrite run it on all disk. Are all HD set to master or single drive. All all HD the same capacity and mfg? I would also monitor the voltage on the power supply, 12vdc and the 5vdc. Weak PS can cause havic.

ALSO The Snap OS is NOT COMPATIABLE with VISTA. DO NOT CONNECT WITH A VISTA PC, CAUSES NOTHING BUT TROUBLE.

[Fredo]

Originally Posted by blue68f100

Have you read the notice on the 4100, if so does your 4100 have the mod or newer rev MB.

The machine(s) were upgraded at the factory. There is a (handwritten) sticker on the MB which says "-003 A", so I am pretty confident that the boards are fine.

If you have the Snap OS do a re-install.

See what I can come up with.

If you have spinrite run it on all disk.

No, don't have spinrite. But these discs are brand new, and we got the exact same results with the "original" discs and "originals" from the other unit. So I highly doubt that this is a disc problem.

Are all HD set to master or single drive.

All set to master.

All all HD the same capacity and mfg?

Yup.

I would also monitor the voltage on the power supply, 12vdc and the 5vdc. Weak PS can cause havic.

That is a good idea.

ALSO The Snap OS is NOT COMPATIABLE with VISTA.

All XP Pro 32 SP3

Will report back after doing more testing.

Thanks
Fredo

[Fredo]

It looks like I had a failing Power Supply. Rebuild of backup disc is at 25% and share/folders are accessible. So I think I am on my way to success.

A question about Spinrite ... This looks like a great tool; worth every cent.
But how do you use Spinrite on a Snap Server? Obviously there is no disc, floppy or USB stick to boot from. Does that mean that we have to remove each and every disc individually and put them into another computer to be able to run Spinrite on the disc?

Thanks for your help.
Fredo

[blue68f100]

You have to remove the HD and install it in a pc. Since Spinrite works at the controller level the OS does not come into play. Just DO NOT ALLOW A PC TRY TO MOUNT THE SNAP HD'S. Doing so will make the HD useless for the Snap.

I use it on every new HD and at least once a year to get an idea as to how the HD is.

[Fredo]

Great success, all went fine and my 4100 is back in good shape.
Thanks for the help and hints.

This leads me to another question.
I still have the 4 discs which were in the unit whent it broke down.
Now that it turns out not to be a HD failure, but a Power Supply problem, I am having a go at trying to bring those discs to life.

Just for facts, the story:

-At one point the 4100 stalled and reported a missing disc.
(Didn't know that it was a Power Supply failure back then)
-We have done a rebuild and apparently everything went fine.

The current situation is as follow:
-When rebuilding, the unit stalls at 75%:

================================================== ==================
08/14/2008 9:49:33 FATAL ERROR PANIC : General Protection Fault (#13) at $00243E93
EAX=FFF09FEC EBX=08000000 ECX=9DEC3FBF EDX=00000000
ESP=003F604C EBP=003F6884 ESI=07F72FB4 EDI=003F7698
================================================== ==================

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

08/14/2008 9:49:32 ERROR File System Check : Cannot Read: Blk -16
08/14/2008 9:49:32 ERROR File System Check : Cannot Read: Blk -16
08/14/2008 9:49:27 ERROR File System Check : Cannot Read: Blk -16
08/14/2008 9:49:27 ERROR File System Check : Cannot Read: Blk -16
08/14/2008 9:49:25 ERROR File System Check : Cannot Read: Blk -16




The disc status reports an Orphan disc 3.
Now that it turns out that disc 3 wasn't failing at all, isn't there any possibility that we can "trick" disc 3 back into the raid, so the system can (hipefully) rebuild itself in a proper way. 'Cause apparently there is now something wrong at disc 1, due to the rebuild with teh failing power supply.
I have ran Spinrite on all 4 discs and they are OK.
Here's the info device log:



Logical Device: 10006 Position: 0 JBOD Size (KB): 32296 Free (KB): 31808 Private Mounted
Label:Private Contains system files only
Unique Id: 0x52A9A02F7E20B2FF Mount: /priv Index: 12 Order: 0
Partition: 10006 Physical: 10007 FS Size (KB): 32768 Starting Blk: 515 Private
Physical: 10007 Drive Slot: 0 IDE Size (KB): 134217216 Fixed

Logical Device: 1000E Position: 0 JBOD Size (KB): 32296 Free (KB): 23280 Private Mounted
Label:Private Contains system files only
Unique Id: 0x10F74A8674968F0C Mount: /pri2 Index: 13 Order: 1
Partition: 1000E Physical: 1000F FS Size (KB): 32768 Starting Blk: 515 Private
Physical: 1000F Drive Slot: 1 IDE Size (KB): 97685504 Fixed

Logical Device: 10010 Position: 1 ORPHAN Size (KB): 97028944 Free (KB): 0 Public Unmounted
Label: Drive3 Orphan from TEMPLE_RS01 - RAID5
Unique Id: 0x0353C49F724C4D08
Partition: 10010 Physical: 10017 ORPHAN Size (KB): 97028944 Starting Blk: 104776 Public
Physical: 10017 Drive Slot: 2 IDE Size (KB): 134217216 Fixed

Logical Device: 60000 Position: 2 RAID_CRACKED Size (KB): 291086832 Free (KB): 0 Public Unmounted
Label:RAID5 Large data protection disk
Unique Id: 0x0353C49F724C4D08 Mount: /0 Index: 0 Order: 255
Partition: 10000 Physical: 10007 R 60000 Size (KB): 97028944 Starting Blk: 104776 Public
Physical: 10007 Drive Slot: 0 IDE Size (KB): 134217216 Fixed
Partition: 10008 Physical: 1000F R 60000 Size (KB): 97028944 Starting Blk: 81942 Public
Physical: 1000F Drive Slot: 1 IDE Size (KB): 97685504 Fixed
Partition: 10018 Physical: 1001F R 60000 Size (KB): 97028944 Starting Blk: 81942 Public
Physical: 1001F Drive Slot: 3 IDE Size (KB): 97685504 Fixed



Thanks for the help support.
(Due to the takeover by Overland, support refuses to talk to me, unless I pay in advance for -maybe- getting support)

Fredo

[blue68f100]

The snap stores drive info in flashram, so if the set is in flash, 2 drives and your out apply here. I know of no way to bypass this.

With drive 1 having a different starting point your in big trouble. The snap use drive 1 (10000) to configure the raid size. This needs to be the smallest drive. Also if you have done software/firmware updates v2 to v3 calculated the starting point differently. So you really needed to wipe all HD's and let the Snap do the HD prep/config like it was a brand new machine with the smallest HD in position 1.

[Fredo]

I see.
It was worth to try though.
The system is working fine now with 4 brand new discs.
I even *think* I have more capacity now (385,412 MB).

Thank you for all the help; I appreciate it very much.

Best regards
Fredo