Snap 2200 not holding o/s

[fisherscpa]

I had a hard drive fail. After replacing the drive (as I have found the instructions in several places here- thank you very much) the system will recognize the drive but have to do some sort of drive check everytime it boots. Additionally, if left on for several hours the unit's just starts to rapidly flash and I have to reboot.

Snap Server Window Reads...
Model- 2000 series
Software- 4.0.855 (US)
Hardware- 2.0.3
Server #- 801089
Bios- 3.4.758

The server log reads...

S
System Initialization : Server v4.0.855
Build Date: Sep 24 2004 20:13:34
Boot Count: 14
System 8/1/2007 11:48:15 AM
F
Call Stack :
$0024B451
$00246004
$002440D6
$00243FA6
$0023C65B
$002399FF
$001041C3
================================
System 8/1/2007 11:36:44 AM
F PANIC : ASSERT(fcbp->fcb_magic == FCB_MAGIC) System 8/1/2007 11:36:44 AM

I have tried using Digital Dolly to reload the BIN file and the process is just repeated.

Alright you great brains out there, help me out. I have spent so much time on this now- it's just the principle of the thing at this point.

I have another unit (2000) that runs just fine without any issues at all. Would installing 1 of the drives from the 2200 and formatting it in the 2000 help?

Thanks,

Jeff

[blue68f100]

A little more specific detail for me to know what you did.
Did you replace the HD with one that is identical (same mfg, model size & rev) or < > in capacity? There are several procedures scattered all over this site. Which one did you use? Did you remember to change the jumper to master or cs depeding on what you original was.

The rapid flashing is panic mode, System error/crash.

You said you did a bin/img file load. Were you able to do the /init command or did it fail to get that far? And did you do it with 1 drive or 2 installed. The fcb error has to do with the XFS file system. Which makes me wonder if the drive is good/compatiable with your system.

[fisherscpa]

A drive on the Mirrored unit failed. I did not know the anatomy of the unit but it turns out it was the 1st or main drive. The error on the debug page read something like "broken mirror". Following the instruction on replacing a dead drive on the unit did not work initially. Fortunately, I did have an image copy of the o/s and the information on the original drive was already backed-up on another drive. My attempts to reint the original drive failed as the debug page still listed the original (and remaining drive) as drive 50000 not 10000.

I reformatted the drive in a in windows system and copied the image file over to the drive using dd.

Once the drive is installed it reformats and initialized fine. It is not until a second (identical drive) is installed, reformated and initializes that I have an issue. Sometimes the issue is immediate. Sometimes it takes a few hours with the system running normally before the rapid system light (panic) occurs. I have not tried to mirror the drives or set as striping. The drives are just 2 independent drives in the unit. Each with a different share name. I have even tried copying the original good drive (the entire drive) to the replacement drive using dd. Again its OK for a while (hours) then the panic error occurs. Sometimes I can merely reboot the system and it remains stable for a couple of hours but eventually I get back to the panic error and unit will fail to reboot correctly.

To answer your questions...
1) Did you replace the HD with one that is identical (same mfg, model size & rev) or < > in capacity?= YES
2) There are several procedures scattered all over this site. Which one did you use? ALL OF THEM EVENTUALLY. I DON'T KNOW WHICH ONE FIRST.
3) Did you remember to change the jumper to master or cs depeding on what you original was. ORIGINAL WAS MASTER BUT THE SYSTEM WILL NOT RECOGNIZE THE 2ND DRIVE UNLESS IT'S JUMPER IS SET AT SLAVE FOR SOME REASON.
4) You said you did a bin/img file load. Were you able to do the /init command or did it fail to get that far? YES (SEE ABOVE)
5) And did you do it with 1 drive or 2 installed. I HAVE TRIED ALMOST ALL THE COMBINATIONS AVAILABLE. I HAVE TRIED INITIALIZING THE DRIVES INDEPENDENTLY THEN PUTTING BOTH DRIVES IN THE UNIT AT THE SAME TIME- NO SUCCESS.
6) The fcb error has to do with the XFS file system. Which makes me wonder if the drive is good/compatiable with your system. THE REPLACEMENT DRIVE IS IDENTICAL TO THE ORIGINAL AND MIRROR. I HAVE ALSO TRIED OTHER DRIVES. ALL WITHOUT SUCCESS.

I have also tried to upgrade the o/s once the drive was operating correctly. The upgrades are reported to have taken place but the system doesn't show the new s/w version which I get a specs page.

Thanks,

Jeff

[fisherscpa]

I have the following bin files...

snap1000.bin
snap2200.bin
2200.3.4.805.bin

and SUP files for the o/s of...
3.4.803
3.4.805
4.0.829
4.0.855
4.0.860

I don't know if that info helps...

[blue68f100]

50000 refers to a mirror RAID1 share. 10000 & 10010 are the actual drive ID's. Normally on a 2200 the top drive is 10000.

Now you say M/S is the only way the second drive will be reconized. This configuration refers to a 2000 model and not 2200 unless you are using Maxtor HD's. The 2000 has the drives side my side where the 2200 has the drives on top of each other. You can NOT mix CS and M/S. Must be one or the other. If you are using Maxtors drives on a 2200 the setting must be single drive or CS. I think CS worked better.

A few things to check. Make sure the system fan is working. The 2000v1 and 2200 both have a system fan. The 2000v2 may have a cpu fan. The CPU is normally temp controlled. I would recommend wiring it so it runs 24/7. Re-seat the ram, to clean up the contacts.

First do a reset to factory settings, if you have a copy of Assist. Check to see what OS version you are running by clicking on the SnapAppliance logo on the admin page. You have said it has not changed. V3.4 on up can be upgraded from the GUI. Must have a lot of paitence, at times it will act like it is hung. If you have your computer to go to sleep quickly, make sure you move the mouse a little every 30 sec or so. If this fails you can also upgrade it using assist. I and other have discovered that Assist seams to do a better job of upgrading, if you suspect you have corruption. If you are using Vista, throw it out the door, not compatable with these older units.

I would only work with 1 drive, top or master if a 2000. I would do a clean install using a image file. Init the drive. With it up, look at the Drive configuration. See if it's reporting drives missing/ mirror broken. If so you must use a debug cmd to reset it. Adding drives or mixing position can corrupt it's settings. If everything is good (drive cap) and OS v4. Install the second hd that has had a quick erase using the MFG utility. Power up and let the snap init the drive. Run "co de info" to verify that the starting position is the same, and the capacity is the same. MFG constanly update firmware, so HD with the same PN can have a different capacity. Make sure the capacity is correct. If OK you can now change the drive config to Raid1. If you have it out of the case, make sure the fans are running. If not it will over heat and generate errors.

I use SpinRite on all new HD's to locate and map out bad sectors. This takes time to run depending on capacity. New drives have not had the media checked, mfg rely on SMART to do it on the fly.

Since DD use a RAW copy, coping 1 drive to another may or may not be 100% correct. Particularly if it had some data corruption. If you do this run SpinRite to verify the data.

[fisherscpa]

1) New WD (160GB) drive set as primary (sole drive)
2) Jumper at Master
3) Factory reset= done
4) New BIN image file copied to WD with DD prior to installation
5) System Boot and config debug command co de format 10000 /reinit= completed
6) System reboot without issue and drive mounts
6.5) Took drive out and did a SpinRite check on the drive= no issues
6.75) System reboots just fine
7) System runs for a couple of hours then PANIC mode sets in
8) Fan is always running
9) Once PANIC starts then the drive always does a drive check and PANICS again at reboot.
10) It doesn't matter the HD put in WD (160), Seagate(120), Maxtor(75) all have the same outcome.

I don't know if this is an OS issue, HD issue or a unit (2200) issue. Any suggestions?

Jeff

[blue68f100]

Sounds like a bad system board, hardware failure. You have eliminated all other posibilities.

[fisherscpa]

Is it worth it to replace the system board? Where does one locate one? Is there something I could have done to cause the board failure? O/S updates in an inproper way...? I have a 2000 as well and don't want to repeat the mistake while trying to upgrade its 30GB drives to larger sizes...

Thanks,

Jeff

[fisherscpa]

Is there a way to test the board?

[blue68f100]

You can buy a complete used unit for what the MB sell for. http://www.northwesttechnical.com/ Sell parts for these.

[fisherscpa]

I don't know why I didn't consider this sooner...

I have a functioning 2000 with smaller drives in it. Couldn't I determine whether its the board or a drive setup by putting the 2000 drives in the 2200? The drives in the 2000 are WD. Would the setting be Master or C/S on the jumpers in the 2200? It appears as if the drives are C/S in the 2000 but the original configuration of the Maxtor drives in the 2200 was Master.

Is there anything else I didn't think of? I don't want to have 2 door stops when I get done with this experiment...

Thanks.