help me hack into, upgrade Snap 2000

[sleon]

I have an old Snap 2000 with 60GB and I believe version 3.x of the OS. However, I can't confirm that, because at least one drive is bad and the snap won't hold a configuration, nor will it let me reconfigure it.

I've scanned the various threads here. Various hints are scattered here, there, somewhere else. I'd like to collect all the instructions in one thread -- here, please. And, I'm convinced I need to crack open the case, rip out the drives, and install new drives.

So, here are my questions:

1. Exactly how do I pry open the hardshell? There are no screws. I've tried wedging a screwdriver, but don't want to break some flange or damage some circuit board. And I'm not an expert at Rubik's cube.

2. What is the max capacity/largest drive I can install? I've seen references to 120, 133 as the limits, but some black-diamond experts appear to have installed 300+?

3. What type drives? Recommended makers and models?

4. What OS version and/or upgrades do I need? And, since the current snap maker seems uninterested in making these available, where do we get these OS versions and upgrades?

5. What else do I need to know and/or do? I'm not interested in the secret of life (I figured that out already -- get up in the morning, go to work, eat dinner, kiss the wife, play with the kid, walk the dog, go to bed, start all over again). I just want to save the snap from becoming a doorstop.

In advance, thanks.

[re3dyb0y]

1. Can't Help - David will be able to when he comes on

2. 137Gb is the key number, Depends on the hardware. I believe, i think you could be alright

3. Seagate or Western Digital

4. You dont need snap OS v4, but it does have some advantages. We can help



Are you actually able to get onto the snap using assist or the web interface?

[blue68f100]

Many questions.

1. There is one screw on the back by the fan. Then the case slide ft to back about 1/2" then lift. The 1/2 with the led should be down.

2. It all depends on the BIOS as for max size. But some have got it to work with the old v1.2.180 BIOS. I only had 80 gigs in mine. The biggest problem is HEAT. You will need to replace the fan to move more air if you install 7200rpm drives, org are 5400.

3. If you look a the cables, if they are 40 cond you only need a ATA 66 interface. I recommend Seagates or Hitachi drives. You want a IDE interface, these onld units do not play with EIDE (some Maxtors and WD).

4. This unit will aut format and setup the new drives. Master is end of Cable (ft) where the slave is closest to the power supply.

5. Since OS is stored in flash ram Stay with v3. Some (w/SN <28000) have had problems with v4.0.860

If you check the wiki section you will find the pdf manuals and Quick start guide. The reset instruction are there also. The only software you need is assist, if you clear the network settings.

You can connect to this unit with out any drives installed. the default name is SNAP(serial number). Web connect is "http://(Snapipaddress)" or do a search use the name.

This should get you started.

[sleon]

You guys are GOOD. Thank you!

Question 1 is now definitively answered.

To pry the case open, unscrew the one screw near the fan. Turn the case on its side, LED side down. Push what is now the top side of case back toward the fan. The case top will pop off.

I presume this will work in reverse when it's time to re-assemble.

Two screws hold the drive cage to the lower case. One screw is near what would be the inside of the top of the LED tower. Look for the second screw about 3 inches straight back from the power socket. Unscrew these and you can flip back the drive cage to get at the four screws that mount each drive.

I have not yet unmounted the drives, nor removed the data and power cables from the drives. I'm planning to do that after I see more answers as to what the replacement options are for new drives.

Anyone want pictures? Am I able to post/attach pix?

[re3dyb0y]

They are useful for reference

I believe you can attach pics

[tc]

I have a used Snap 2000. It is branded Quantum and has dual 10gig's inside. 2 IDE channels both drives are set to master. I have put a 128MB dimm into the machine.

The software : Software Hardware Serial# BIOS
2.2.393 2.0.0 35236 2.0.252

what options are thre for upgrading this unit?
I have a 250gig western digital drive and a 120gig seagate
the western digital comes up as a 32gig when I install it

Any help appreciated

[Parity Error]

2000 sn 51xxx
Product: 0 - "Minnie: 2 drive IDE" - Platform Bytes "2.0.0"
Server-3.4.803, Build Date: Jan 15 2003 18:04:19
{where do I dump HW/BIOS versions?}

RAID active

Went to replace the one dead 40GB drive with another. (Right Hand drive is DOA). But a previously working IBM Deskstar I installed, with identical CHS, had lots of head-banging as if it were dead, then after a power cycle or two, all was well.

?Bad Omen?

Now I have one empty Drive 1, and one Data Protection Drive (0, I guess?).

I want to reestablish the RAID, but DO NOT CARE about the data that's there. Do I just reformat the so-called Data Protection Drive or.....?

[Parity Error]

2000 series
3.4.803 (US)
2.0.0
2.0.282

I have read the wiki on restoring a mirror, but have one important question. "info devices" reports:


Logical Device: 10006 Position: 0 JBOD Size (KB): 32296 Free (KB): 22576 Private Mounted
Label:Private Contains system files only
Unique Id: 0x1B24155171517FE2 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): 40208896 Fixed

Logical Device: 10016 Position: 0 JBOD Size (KB): 32296 Free (KB): 23784 Private Mounted
Label:Private Contains system files only
Unique Id: 0x71308F80436544F0 Mount: /pri2 Index: 13 Order: 1
Partition: 10016 Physical: 10017 FS Size (KB): 32768 Starting Blk: 515 Private
Physical: 10017 Drive Slot: 2 IDE Size (KB): 39088640 Fixed

Logical Device: 10000 Position: 1 JBOD Size (KB): 39359160 Free (KB): 39319328 Public Mounted
Label:Drive1 Single disk
Unique Id: 0x1B24155171517FE2 Mount: /0 Index: 0 Order: 2
Partition: 10000 Physical: 10007 FS Size (KB): 39904920 Starting Blk: 37869 Public
Physical: 10007 Drive Slot: 0 IDE Size (KB): 40208896 Fixed

Logical Device: 50000 Position: 2 MIRROR_CRACKED Size (KB): 38259736 Free (KB): 16564752 Public Mounted
Label:Mirror Data protection disk
Unique Id: 0x673D5C6E31D4D57E Mount: /2 Index: 2 Order: 3
Partition: 10010 Physical: 10017 M 50000 Size (KB): 38790264 Starting Blk: 37169 Public
Physical: 10017 Drive Slot: 2 IDE Size (KB): 39088640 Fixed


The wiki says:

4. After the ORPHANED disk is formatted, execute this command from the ipaddress/config/debug page.

* co de config mirror 100SS 100DD
* SS is the source disk, DD is the destination disk.
* (This step mirrors the drive with data to the drive without data.)

BUT are SS & DD the physical or the logical or the partition labels?
Do I want
co de config mirror 10017 10007
or
co de config mirror 50000 10000


Since 10017 is both the ID on the one drive AND on the one flash device with the OS...I really do NOT want the wrong one here.....

[blue68f100]

Back up any critical data, before doing any thing.
Always safe than sorry.

It should be the physical ID of the drive. co de config mirror 10000 10010

If you look at the drive info the new drive should be mounted, showing a free share.

10017 & 10007 are partion used by the OS, where 50000 is the raid1 set.

Please this will only work if the drives are the same size. If not let me not there is another work around.

[Parity Error]

Always save than sorry.


Good line; but in this case, I do not want the mirrored data, I just want to reestablish the RAID for new data...

It should be the physical ID of the drive. co de config mirror 10000 10010

But the top two entries have the same physical ID as the 3rd & 4th. Are those partitions on the drives; I'd thought they were the NVRAM that held the OS. [I'd assumed that the 2000 was a model with OS_on_Flash, not OS_on_disk -- true or false??] If so, the format I did on the replacement drive had to created the one...

If you look at the drive info the new drive should be mounted, showing a free share.

The GUI shows Drive 1/Single Disk/OK


10017 & 10007 are partion used by the OS, where 50000 is the raid1 set.

Please this will only work if the drives are the same size. If not let me not there is another work around.


My replacement drive is slightly larger; 38,436 vs 37,363 MB. With software RAID on Linux, this was not an issue -- is it a problem here?

[blue68f100]

The 2000 store the OS in flashram. The OS creates serveral slices for keeping directory info/permisions cache .... Std UNIX file structure.

If you are wanting raid 1, go a head and issue the mirror repair. You can just blow away any data left.

I should size to the smallest drive. Snap says it has to be with in 2-3% which you are.

If that does not work try...

From the disk utility change to JBOD (single drives).

Once this is completed change it what ever raid or disk setup you want.

If that does not work we will have to issue a force cmd to break the raid.

[Parity Error]

The 2000 store the OS in flashram. The OS creates serveral slices for keeping directory info/permisions cache .... Std UNIX file structure.

good to know

If you are wanting raid 1, go ahead and issue the mirror repair. You can just blow away any data left.

Not sure what command this is..... the GUI Config Disks is locked down; can not change. Do you mean a 'config dev con' of some ilk?


If that does not work try...

From the disk utility change to JBOD (single drives).

Once this is completed change it what ever raid or disk setup you want.

No joy

If that does not work we will have to issue a force cmd to break the raid.


Another point I'm confused about; you previously said I should use the PHYSICAL addresses, but then went on to say
co de config mirror 10000 10010

but those descriptors are Partition, not Physical..

[blue68f100]

The 10000, and 10010 are the physical address of the drive. "co de info" will reveal this.

If the above does not work try "co de config individual /force" It's been awhile since I've used this so it may not be correct. Try variations like "co de config individual 10000 /force" if all goes correctly it will restart an re config the drive.

This need to be changed to get full control of the drive.

Or use dd and copy drive 10000 to 10010.

[Parity Error]

The 10000, and 10010 are the physical address of the drive. "co de info" will reveal this.


I posted mine above; here's the short form:

Partition: 10000 Physical: 10007 FS Size (KB): 39904920 Starting Blk: 37869 Public
Physical: 10007 Drive Slot: 0 IDE Size (KB): 40208896 Fixed

Partition: 10010 Physical: 10017 M 50000 Size (KB): 38790264 Starting Blk: 37169 Public
Physical: 10017 Drive Slot: 2 IDE Size (KB): 39088640 Fixed

Hence my confusion over Physical vs Partition....

[blue68f100]

Partition: 10000 Physical: 10007 FS Size (KB): 39904920 Starting Blk: 37869 Public

if you look closely the 10007 is inside the 10000
The xxxx7 is the slice used for data, Yes it is a little confusing. I thought the same thing till a Snap tech helped me out. Which is where the repair an Orphan Drive came from.

[Parity Error]

Well....
co de config mirror 10000 10010
got me an Error 3
whereas
co de config mirror 10010 10000
got an Error 4


I did BOTH drives with cd de individual commands and THAT unlocked the disk configuration option; so I could build a new RAID...

And the update just finished....

So it looks like about everything is OK, with one exception.

Thanks to all for the help..

BUT I have lost the SSL protection on the web interface, and can't see how to reenable it.

[blue68f100]

With these snaps their are a lot of options. With the 2X00 series in raid 1, as long as the main (10000) drive is good its a easy. Some times you have to rotate the slave (10010) into the 10000 positions.

SSL is handled through the JVM, have not used it, sorry.

Sam or one of the other should be able to help you.

[re3dyb0y]

You need the JVM installed and running i believe

It has the SSL files as david said

[Parity Error]

You need the JVM installed and running i believe

But I seem to have lost same in the course of rebuilding the mirror. I don't have the original CD's here but can get them -- is the JVM on one of them?


I found a new problem as well; Apple Networking is dead. SMB works on the Mac, but not AFP. It worked before I restored the mirror.

Further, I can not get SMB working from a W2K box. Color me confused...

[re3dyb0y]

Are you sure that the network settings for those haven't been turned off?

If so, turn them on and reconfigure

Or remove and re-enter if they are



I dont know if they are on the cd's

I have it somewhere

[Parity Error]

Are you sure that the network settings for those haven't been turned off?

If so, turn them on and reconfigure

Or remove and re-enter if they are


They are on. I turned them off, and back on [rebooting in between] to no avail.

Bit it looks more complex than I thought; and I've screwed things up again...

smb//172.16.73.222 works on the OSX box.
afp://172.16.73.222 works on the OSX box.
But browsing does not show me the SNAP.

Further, I screwed up and told my Mac to demand encrypted passwords; now I can't see how to remove that [OPTIONS is grayed out..] And the SNAP lacks that.

On the M$ Box, I'm still nowhere..

[re3dyb0y]

Thats a strange IP range to use

[jontz]

Thats a strange IP range to use

Not really. The IANA has reserved three blocks of addresses for private use.

10.x.x.x as a class A
172.16.x.x as a class B
192.168.0.x as a class C

Evidentally he needed more addresses than a class C could provide, so he used the class B private range instead.

[re3dyb0y]

Oh yeah

Thought it looked familiar


So why do places use Class C and then use .10.x and such?



I know of a company that used the Class A and then the first x beign the place (world wide company), 2nd being printer, server etc, and the 3rd being the actual item)

Mind enlightening me on the difference between the classes?

[jontz]

The main reason that you would use a higher class of IP would be so that you could more easily subnet off a range of IP addresses. For instance, a customer of mine recently installed about 40 IP cameras in 5 different buildings. They put all of the cameras on 172.16.12.x. By having all of the cameras in the same 12.x range it was easier for them to group devices together. It also made it easier for subnetting. They could subnet off the entire 12.x range of IP's so that normal users couldn't see the cameras from their workstations since all of the workstations were on 172.16.10.x. It's much easier than using a class C and figuring out different netmasks for different users.

The other main reason to use a higher IP class is that you need more IP addresses. The only real difference between the IP classes is the number of IPs in that class. Class C has 254 usable IP's, where a class B has 65,534 usable IP's, and a class A has almost 17 million. You could set up a 10.x.x.x. for your home network if you really wanted to. When you have 17 million computers in your house, let me know and I will fly to the UK personally to see it :)

[re3dyb0y]

When you have 17 million computers in your house, let me know and I will fly to the UK personally to see it :)

Well

I could get machines with many network cards...

And other devices...

But its still gonna take ages

:P

[Parity Error]

Not really. The IANA has reserved three blocks of addresses for private use.

10.x.x.x as a class A
172.16.x.x as a class B
192.168.0.x as a class C

Evidentally he needed more addresses than a class C could provide, so he used the class B private range instead.


No, I just wanted to be in a separate block. And there's a 1-1 mapping with public addresses: i.e. 172.16.73.220 maps to xxx.xx.73.220, etc -- it's easier for my feeble brane that way..