Hack a Snap Sever 1000?

[rickytim]

No I can get on to the web interface with the hard drive I used the DD command to. When I go to disk management and go and try any of the options that come up I get a error saying no drives where found.

[blue68f100]

NO ? I CAN ? Double negative, OK

If I read this correctly you can login through the web interface. But the new HD is not visable.

Did you use a WD or one of the maxtor drives with a ATA133 interface? Snaps prefer ATA100. Re check the jumpers on the drive.

If you have Partition Magic pull the drive, delete any partitions found. I found a quick fat32 format works good for this.

Since the first attemp didn't work. I will send you the instructions to copy the boot tracks from your original drive.

[rickytim]

The original drive does not work. I messed it up when copying the image over to my new hard drive.. So I do not think I will be able to use it for boot tracks..

The hard drive I am useing is a WD model numberwd2000jb. I can not find if it is a ata133 or ata100. And I made sure my jumpers are set to master when hooked up to the snap server..

[blue68f100]

I checked the model, its a ATA 100.
Was your original drive set to Master or Cable Select.

Does the 10 gig drive show up, or is it missing in action like the new one. You pevious posted that the new drive was not found.

[rickytim]

The original hard drive was set to master. When I try the 10gig hard drive in the snap it does nothing I can not login to it at all. The only way I can get to the web interface is by using the DD command. I have tried formatting the hard drive with partition magic and trying it on the snap but would not work either. All that happens is the 1st and the 4th led blink at the same pace in sequence with both the original 10gig hard drive and with a fresh format to fat32.

[blue68f100]

With both lights in unison means? the os is missing? If the disk light is solid with any of the hard drives, with system blinking some what, mainly on. This ususally means it is setting up the hd. If this is happening you will have to wait 10 minutes before you can access it. I am wondering if you over wrote the image file. If you have the images files in the zip. I would un zip and try again. One thing to remember. With the XFS file system, it will not overwrite if it sees any nfs or XPS boot tracks. Do you have another ide cable to try. I know these are on there very tight and could have been damaged during removal.

Dont be afraid to reset to factory settings if nothing happens. Some times it needs a boot.

I just had a 1000 delivered to me yesterday, to try to fix. He was playing with turning a 1000 to a 2000, when the snap rebooted before he got something set. He says he has no interface, web or assist, MIA.

It want be till friday before I can do any thing else.

Have a Happy Thanksgiving.

David

[Marvin_pa]

Hi all, glad to see this thread is still going stong. It's been a long time between posts for me

I have obtained a couple more Snap 1000 drives and have from what I read it seems you can break through the 137gb barrier on them ?

Is there any specific requirements of the hardware ?

The one unit i have to hand has the following spec's -
h/w 3.1.2
s/w 4.0.885 (just upgraded the other day)
bios 2.1.336

Can I install a > 137gb drive in this ? Is the process much the same as a < 137gb drive ?

Thanks all,

[blue68f100]

Yes you can break the 137gig. Your Software & hardware meets the minimum requirments. The easiest way is to use dd from the knoppix cd to copy your boot tracks (and data if you want) from your existing drive to the new one. Then use debug and issue "co de format 10000 /reinit". On reboot the snap will init the drive and gain full capacity.

[jrmvt]

OK, I have a 705N which I FINALLY figured out had a bad hard drive (lights on the front were no help on this one). Replaced the Hitachi 40GB drive with an 80GB of the same line (smallest I could find). Could not get it to rebuild the RAID 5 Array , so since I had a good backup of the data, I just broke the rest of the drives apart and started from scratch.

Well, I figured this would be a good time to upgrade the OS to 4. Followed the directions to fool it into thinking its a Snap server and upgraded to 4. Appeared to go off without a hitch. All 4 drives ran through the disk check with no problems while broken apart. I built the RAID 5 array and still no problems.

Then I ran disk check on the RAID 5 disk and it blew up. I get an "FSCK fatal error = 39" and "Bad State 0 for I=237628". I did a complete reset to factory settings (all 4 options) and broke apart the RAID again and all disks checked out individually. Built RAID 5 and all OK. Ran Disk Check and it blew up again.

Anyone have any ideas? I'm afraid its something messed up in the BIOS. I don't need larger drive support as its not storing that much data, but I do need Win2003 server support. That's why I moved to OS4.

[blue68f100]

There has been some reports of the 4000 having problems with large drives. Being a dell, I was suprise you didn't use dell's down load section for the guardian file. I think v 2.6 gives you the Win2003 server support. 40 to 80 gigs should have not been a problem. Raids are can be very picky with hardrives. It is allways best to have all drives at the the same hardware and firmware versions. If you are using the JVM, you may need to increase your ram. I recall jvm requires a minimum of 24 meg. The 2003 server support is interfaced with java script.

The error is due to a parity checksum error. There have been several users that just gave up on raid5 after upgrading to hd>137gig. When you do checks like this, it also stress the ram. The 4X00 experts (misterBill & poogles_uk) may be able to shine more light on the problem.

[JPB]

Originally Posted by phaedrus

I think this is a hardware limitation - probably the Promise controller and/or its BIOS.

I have a 4100 which started life as a Dell 705n. It's running 3.4.805 which is 48bitLBA compatible unlike 3.4.803 which is not. 4.0.830 is also 48bitLBA capable, but only on suitable hardware.

Which is not really true. If you "bios pci", you get the vendor and product ID's.
Devices on the PCI bus:
# 0 0 0x8086 0x7100 *430 TX Chipset memory controler
# 0 7 0x8086 0x7110 *82371AB (PIIX4) PCI-to-ISA Bridge (fcn 0)
# 0 9 0x8086 0x1209 Intel
# 0 10 0x105A 0x0D30
# 0 11 0x105A 0x0D30
The last 2 are the IDE-Controlers, which has just the same Vendor and Product-IDs as Promise FastTrak 100. Earlier FastTrack had problems with 48bit LBA, but flashing the cards with a newer BIOS resolved the problem.
This is still no answer on how to enable 48bit LBA on 4100 with to old BIOS, but I see no reason why it should not work. There are 2 ways updating a Snap-server:
1. With a newer OS - normally resides in a flashrom
2. with a newer BIOS file with either TFTP or bootp (bios Flup [T=>TFTP, B=>Broadcast])
I also get
11/29/2005 9:41:49 98 D SYS | IDE: A 48 bit LBA capable Hard Drive was detected ...
11/29/2005 9:41:49 98 D SYS | IDE: 48 bit LBA is not supported on 4100 and 12000 platforms.
on my 4100's here, but I guess that's not true by hardware, but true by Adaptec. If anyone get connection to
11/29/2005 9:41:49 98 D SYS | Executable built by KEVIN
he could tell us how to flash a newer BIOS. Another approuch is to extract the BIOS from another 4100 with working 48bit LBA. Setting up a bootp-Server is not really complicated and TFTP is a well known service. We still don't know, if the BIOS file needs to be packed or altered in any way to get flashed.
Last chance: desolder the controler-chips, put them on a FastTrack and update their BIOS in a PC... ^^
There are several hints how to hack onboard Promise chips on ASUS and other motherboards. Promise uses the same chip for standard IDE and for RAID 0/1-devices. You can extract the Promise-BIOS from the motherboard-BIOS and include the other BIOS to change the Promise into something else (the device ID will change - beware!). All we need is a 4100-BIOS with or without working 48bit-LBA and some time to find the Promise-code inside the BIOS. And a hint on how the file must be provided by TFTP or bootp.

btw.. I wonder if you can boot eg. DSL with bootp.. hmmm...
Guess I gonna play a bit with "bios Flup B"...

[blue68f100]

Some have turned on the 48bitLBA through debug, and have had no problems. Most of the problems were with the older 4000.

Some have used 3 drives with spare without any problems. The 4 drives in array did not work.

[Winter]

Hey all,

I'm trying to get my 4000 past the 127Gb barrier, anyone know where I can find the SnapOS v3.4.805 to download?


Great thread

-------------------------
Software
3.4.803
Hardware
2.2.1
BIOS
2.4.437

[jrmvt]

Originally Posted by blue68f100

There has been some reports of the 4000 having problems with large drives. Being a dell, I was suprise you didn't use dell's down load section for the guardian file. I think v 2.6 gives you the Win2003 server support. 40 to 80 gigs should have not been a problem. Raids are can be very picky with hardrives. It is allways best to have all drives at the the same hardware and firmware versions. If you are using the JVM, you may need to increase your ram. I recall jvm requires a minimum of 24 meg. The 2003 server support is interfaced with java script.

The error is due to a parity checksum error. There have been several users that just gave up on raid5 after upgrading to hd>137gig. When you do checks like this, it also stress the ram. The 4X00 experts (misterBill & poogles_uk) may be able to shine more light on the problem.

OK, my mistake. I said I thought something was messed up in the BIOS but I meant OS. I did not do any upgrading to the BIOS. I was under the impression from reading here and on Dell's forum that OS 4 was the key to being able to have Server 2003 support. I didn't think Dell had updated anything for the 705N after several years ago (before the release of Server 2003). And their website says that the 705N is not compatable with Server 2003. Am I completely off base here?

OK, my drives are not larger than 137GB so that's out. Do you think my next best bet would be to replace the other three drives or increase the RAM? Or am I too concerned about this - should I just build the RAID 5 and use it and not run the disk check routine?

Thanks again!

[blue68f100]

Since you are at v4, it does support Server2003. As for as raid5, I would use 3 drives in array with 1 spare. It should test ok with this config. You will get better performace if you increase the ram. Just use drives that have the ATA100 interface, Not 133.

[AdjeH]

I've read almost all the pages but did not find an answer to my problem. I have a Snapserver 1100 with a harddiskproblem. It makes a terrible noise and is not readable. Can someone make me an image so i can use an other harddisk?
I will pay an reasonable price for it.

[FirebirdTN]

Originally Posted by AdjeH

I've read almost all the pages but did not find an answer to my problem. I have a Snapserver 1100 with a harddiskproblem. It makes a terrible noise and is not readable. Can someone make me an image so i can use an other harddisk?
I will pay an reasonable price for it.

Did you ever get hooked up? I wish I could help you, and if there is ANY way that I can please let me know. Problem is I don't know how to "image", only copy from one drive to another.

I have ZERO Unix/Linux experience, so I tried a couple DOS based cloning methods mentioned EARLY on in the first few pages of this thread.

So far, I like the Norton Ghost BOOT FLOPPY method the best. When I tried that other DOS cloning software, it worked, but after slipping in the newly cloned drive the existing file shares were "X"ed out. Its not a big deal since you have to REINIT the drive to obtain full drive capacity anyway, but it did concern me a little that it obviously didn't make an EXACT copy since the file shares were Xd.

Ghost on the other hand, made a PERFECT copy, with existing file shares intact. I still had to REINIT to get full drive capacity, but I feel more comfortable with this methoed.

Which brings me to your delimna; If I knew the EXACT command with Norton to do a "raw IMAGE" instead of a "raw CLONE", I'd be willing to help you out.

Anyone else use the Ghost method and know the exact command?

-Alan

[FirebirdTN]

Ooookkkkaaaaayyyyyy.....WTF....Are there are other drive size limits besides the 137Gig barrier?

A few pages back I mentioned a successful upgrade from the stock 40gig to a 60Gig WD drive. It was all I had at the time.

Based on that success, and other posts on the 137Gig limit, I figured I'd get a 120Gig drive.

I did the EXACT same process, and now it no - workie....

Actually, the Snap 1000 WILL BOOT FINE with the new drive, however, any page I select in "Disk Utilities" (such as disk status, format, etc) says that NO DRIVES were found!!!

When I goto the debug page and try to force a REINIT of the drive, it tells me bad device ID. If I do a "devices info" command, it spews out ID's all the way down the page, on and on forever, until the Snap is no longer responsive and locks up.

I put the origional 40 Giger back in, and it works just fine, so maybe the Snap just doesn't like my new drive?

Thoughts anyone?

Origional Drive: Quantum Fireball 40Gig. Sucessful upgrade drive was a WD600BB (7200rpm, ATA100, 60Gig WD). Current attemted drive WD1200SB (7200rpm, 120Gig, *ATA100* I think-specs don't get specific, going to check that out in a second, "RE" server series drive.)

OS is 3.4.805, BIOS is 2.1.366, Hardware 3.1.2

Thanks!

-Alan

-EDIT- UPDATE: Got home and tried again with my Snap 1100. The 1100 likes the new WD "RE" server drive, but I tried EVERYTHING and couldn't get it to work in the 1000....Odd....

[AdjeH]

My Snapserver is back online again. Two forummembers send me a link to the image of an 2200. I used Knoppix to "DD" the image to a new harddisk and it worked.

[FirebirdTN]

Originally Posted by AdjeH

My Snapserver is back online again. Two forummembers send me a link to the image of an 2200. I used Knoppix to "DD" the image to a new harddisk and it worked.

Glad to hear your back in business!

My saga continues:

After reading thru TONS of posts (dang this thread is huge; wish it was broken up into its respective separate discussions), I have come to the conclusion that my new WD1200SB is just not compatible with my Snap 1000. Especially helpful was the one post (#200 I believe) where one user reported he could get a 250Gig drive to work fine, but a 120Gig refused to work in his box.

The WD1200SB I have works perfectly in my Snap 1100, just not the 1000.

Just wondering if anyone has compiled a "definitive" list of compatible/incompatible drives for the Snap servers.

I now have a WD1200BB drive on its way in hope that will work.

Only difference between the two drives is 8meg cache vs 2meg cache, build constucion (the RE drive is HEAVY comparitived to the SE drives), and warranty.

It would have been nice if I could have used the "RE" in my 1000 as I wanted to RAID my 1100 with mobile drives, but what a pain in the butt to have some drives work, and others not! And just to be safe, all drives in question are ATA100, NOT ATA133.

-Alan

[blue68f100]

I believe the newer Snaps supports ATA 133 were the older 1000 do not. On large disk it is a hardware restriction if you don't see over 137 gig. Has been tied back to v2 BIOS & Firmware required.

David

[FirebirdTN]

Originally Posted by blue68f100

I believe the newer Snaps supports ATA 133 were the older 1000 do not. On large disk it is a hardware restriction if you don't see over 137 gig. Has been tied back to v2 BIOS & Firmware required.

David

That is the thing though, the new drive I have IS ATA100, and still doesn't work in the 1000. Works just fine in the 1100 though.

-Alan

[mls15000]

I am so glad I found this forum. I just trading an old tape drive for a snap 4000. Did not know they could be upgraded. I have 4 120GB drives I am going to stuff into this, currently has 4 30GB drives. If anyone still has a copy of version 4 OS, I would greatly appreciate it. I run 2003 active directory and only have OS version 3.4.790 (US). Thanks! mls15000@gmail.com

[FirebirdTN]

Update on my Snap 1000 upgrade:

Evidently, there ARE compatibility issues with some drives. Reading a few posts of mine back, I could NOT get a WD1200SB drive to work in my Snap 1000. However, it worked in the Snap 1100 I had JUST FINE!

The drive in question was an ATA100, not ATA133 like some people have warned about earlier on in the forums.

I just picked up another 120 Gig drive, this time a WD1200AB, and it works perfectly in the 1000!!! I am a little miffed, as I picked up the drive on ebay, and the guy advertised it as a WD1200BB (7200rpm), but in fact he shipped an "AB" (5400 rpm). Grrrrr. But it works so I am happy. Much better than the 4500rpm 40 Gig Fireball it replaced.

The moral of my story: If you try this upgrade and follow the instructions found throughout this thread and find it doesn't work, and your positive you did everything correctly, then chances are you might just have a drive that is not compatible with your Snap.

A special "thank you" to the person who reported waaaayyyy early on in this thread that he could not get a 120Gig drive to work, but did get a 250Gig drive to work in his Snap server. And they were from the same drive family. It was his post that gave me a clue.

I hope my experience helps someone out there.

-Alan

-EDIT- 12/22

I just noticed, after the completion of the upgrade, the Snap Server's "Help" files are missing (the files that are associated with the "help" icon in the upper right hand corner of the Admin page). If you are a purist like me, you might want to restore them, even if you never really use it. All you have to do is using the "Assist" application, re-apply whatever OS upgrade your using and they will be reinstalled. In other words, I am using OS 3.4.805. Even though I already use that OS, I re-applied the 3.4.805 update and it restored the help files.

[Dino]

I have a client that has a SnapServer 1100 80GB that seems to have died. I'm new to Snap Servers, never used a working one and just have the Admin guide to go on.

I've read all the posts here and don't see anyone else having the same behavior as the one I have. I hoping the experts here can guide me in resurrecting this unit.

Here's what it's doing:

Assist cannot find the server. Tried with and without a DHCP server (without was just my laptop,a hub and the Snap.)

Performed the one, two, three and four flash resets. I even discovered an undocumented 5th and 6th flash reset (What do those do?)

Still no Assist access.

The Power is flashing once per second.
The HD light does a double flash every second.
Link is solid and Activity led flashes with network usage.


Pulled onboard battery for 2 hours and performed a 4 flash reset, No Access.

The real strange thing is that the Power button does not power the unit up or down (no 3 Power Led flash) even after holding it for 2 or 3 minutes. The only wa to power it down is to pull the power cord.

I puled the HD and ran Maxtor's disk diag on it (read only test, didn't do a low-level format) and it didn't find any problems.

If I understand the postings correctly, a portion of the 1100's O/S resides on the HD, correct?

Does the power down and the ethernet/web access portion of the O/S reside on the HD?

If so, then barring any motherboard issues, I should only need to restore the image back onto the HD, correct?

Thanks in advance for any help you can offer.

Have Happy Holiday

BTW: Adaptec wants $120 up front to open a Tech Support call (refunded and a new unit at no charge if it's a hardware problem). That would be too easy (not to mention I have a funny feeling that they would say it's not a h/w problem and keep my $120)