Upgrade drives on a 410

[Digital Blizzar]

Evening all. Need a little help upgrading the drives in my snap server. I unfortunately assumed that I could just shutdown the server, pull out the 500GB and put in 1TB drives, rebuild the raid and call it a day. Now I'm reading the Guardian OS is on the drives. That's what I get shooting from the hip. Anyway, what's the correct process for upgrading. I read that I should put Drive 0 back in, then put a 1TB in 1 and let it replicate, then swap out 0 and rebuild the raid. Is that right? I don't want to ruin my original drives so any direction would be greatly appreciated. Thanks!

[blue68f100]

If you want to keep the data that's on it you can not upgrade. All data is lost during the operation. There are several post reguarding this on the 4500. The procedure is the same. AV, Snapshots, Diaster recovery, and any other add on service must be terminated, other wise it will prevent the upgrade. If you have the GSU, it even better, but not required.
The basic procedure is this, power down, remove drives 2-4. Install new in 2. Powerup, watch for the OS transfer (5-10min). Power down remove 1, move 2 to 1. boot, add the other deives.

If a service is running, the cold boot will revolk the recovery console. Then you have more options, but require the GSU.

[Digital Blizzar]

I have all of the data backed up already. This was a backup mirror but we're looking to upgrade the drives then make it the primary. Well I was able to use the original drive 1 to boot the server. I installed one of the new drives in the second slot. I then hot swapped drive 1 with a new one and then put in 3 and 4. Everything seemed to be working. I created the shares and mirrored the data again. Everything was fine all week and tonight I was going to just change the IP address on the backup with the new drives to make it the primary and that's when things went south. For some reason, after the network reboot, each of the 5 shares had the same contents. Each folder was exactlty the same. Not sure what to do so I initiated a restart. Well, now it won't come back up. I put the old drive 1 back on and was able to get it back up. Now I've read on here that you should have the new drive in 2 before booting up and that will copy over the OS. Well, the system won't boot if drive 2 is in. But I can add it in after and it's recognized. Anyone have any ideas???

[blue68f100]

The controller on #2 may require drivers to be loaded due to a non bios access. Try slot 3 and see if it works.

[Digital Blizzar]

Ok, I formatted the 1TB drive and put it in slot 3 and it didn't work. First I tried having both in the server at the same time during boot but it would not boot. Then I left the first drive in, booted then put the drive in slot 3 and waited. Shutdown, took 1 out and put 3 in its place and that didn't work either. Not sure which way works or how long I should wait. I waited about 10 minutes each time. What steps should I follow?


What I don't understand is why when I did this originally, the system stayed up when I removed slot 1 and put a new drive in it. I was able to create the raid and everything but it just wouldn't boot when I restarted it.

Anyway, thank you very much for your help with this.

[blue68f100]

I have not worked on the newer hardware from Snap/Adaptec/Overland. Some have been suscess other not.

I do know if there is a file system (formated) it will not use the HD. It must be clean/ no format. You can use the quick erase supplied my the hd mfg cd that comes with the HD's.

Normally if you can get into the bios I will show how many SATA controller it has. And if it's using 2 different chip sets. Some have multiple controller. Those that do, require the linux to install the drivers before they come active.

What have worked on some of the newer units. The procedure I used for 4500 worked for them.

What will fail them is, Raid 5's, AV, SnapShots, and any other program that keep ID' s of the HD. With every thing turned off. remove all HD except #1. Install new into slot 2. then cold boot, you have to watch the HD led on #2. This will vary based on system speed. On the 4500 it take about 6 min on cold boot. Then you have the OS on a new hd. Move to 1 then boot, hot swap remaining HD's. But these are with IDE HD's not SATA.

If you can boot with your original HD in slot 1, do so and hot swap the new HD's 2-4. Then fail drive 1, make sure it is reporting correctly. Once it has been reported, install a hew HD in slot 1 and let it do it thing.

Most of these have a watchdog bios. So if the HD that is suppose to boot and fails after about 10 min it should roll to the next HD.

If you have access to the BIOS I would check some settings for HD ID's of capacity

Adaptec did say years ago they were going to try to prevent users from upgrading the servers. They wanted to be able to charge $500 for a HD you could pick up anywhere for $100. To do this they had to have a setting somewhere in the bios or MB to set a flag, go or no go.

Keep hammering you will find what holding it back. Just start by erasing the HD with the quick erase.

good luck

[Digital Blizzar]

Ok here's what I did next.

1. Put the original 500gb drive back in slot 1.
2. Booted the server and was able to access the guardian OS.
3. Verified the drive was recognized
4. Put the unformated/unpartitioned 1tb drive in slot 2
5. Waited 1 hour
6. Pulled the 500 gb drive from slot 1
7. Put an unformatted/unpartitioned 1tb drive in slot 1
8. Waited 1 hour
9. Restarted the server

After all that, still no dice. One thing I did notice was that when I checked the drive that was in slot 2, it got partitioned and there were logical drives!?!?!

I will be using a RAID 5 but the RAID has already been removed from the original drive so I don't think that's it.

You mentioned checking the BIOS. How would I get to that? The SNAP only has a serial port in the back

Also, I downloaded the disk management tool from Western Digital to format the drive but I'm not sure if that was a good idea or not. Do you recommend another tool for the quick erase or is that or the windows version sufficient?

Again, thanks for all your help with this!

[blue68f100]

Do you have any services running like AV, SnapShots, Backup..... These programs keep up with what HD's are in what slot.

With a null modem cable connected to the serial port you should be able to access the bios and see the actual boot process using a terminal program. This may give you some insight as what is going on. I do not recall what baud rate, thinking 115.2k N81

[Digital Blizzar]

No I don't have any of that running. I was able to connect via the serial cable using 115.2 like you said. Here are the settings in the config

SiI 3114 SataLink Bios Version 5.0.56

1. Boot protocal: Preboot execution environment (PXE)
2. Boot strap type: Auto detect
3. Link speed: AutoNeg
4. Preboot wake on lan: Enable

When I closed that, the system boot but gets stuck on:

0 WDC WD 1001FALS-00J7B

Which is the first hard drive. I left it for an hour an there wasn't any change. Previously you mentioned there may be a flag preventing the upgrade. I didn't see a setting like that in the bios screen. I think this may be the case because it's behaving like it doesn't recognize the hard drive. If it was a matter of the OS not copying, it would theoretically recognize the hard drives but just not boot right? What do you think?

[blue68f100]

Is there a setting in the BIOS to reset to factory setting?

With the null cable setup were you able to watch the complete boot process?

If you have all of the orig hd's. Install them in the order they were removed. Setup a R5 then fail drive 1. May want to do a cold boot in between HD's. Install new let it rebuild once complete do the next and continue through the process till all have been replaced. The delete the R5 reboot, reboot then build a R5.

I have run out of options to try......

[Werner_q]

I just stumbled across the same Problem, unfortunately I completely killed (somehow, don't ask ...) the Guardian OS. Initially I just wanted to upgrade from 4x250GB to 4x1TB, and since I saw the SnapServer 410 beeing sold with that config, I thought, it wouldn't be a Problem.

Ok, so I made the same experiences, Drive swapping, all fine - until reboot. When I successfully had destroyed GuardianOS even on the original Disks , I mailed Overland Storage Support - they told me I had to buy a Warranty Extension ($500,-) before they could send me the OS-Image. I have already payed for that, so w.t.f.; but didn't want to arguee with them.

Next, I decided to try FeeNAS, inserted an PCI Graphics Card, attached USB-Keyboard and USB-CD-ROM. Might be worth to mention, that an ATI Rage IIc didn't work (no Picture after BIOS Post-Messages). I found a stone-old S3Trio64 working perfectly. Probably you don't even need a graphics card when you connect over the serial line. But its more fun (faster !) with a graphics card. Also you have to disable this Watchdog-thing in the BIOS (all other BIOS Options can stay unchanged).

And then I noticed that my (only ?) problem was the SATA-Controller, which stuck shortly after recognizing the first Drive !! And, funny thing, when I removed the Drive(s), rebooted FreeNAS from CD-ROM, re-attched the Drives once FreeNAS had booted, everything worked fine ! The Drives were recognized and working !

OK, BIOS Update for the SATA Controller (which is a SiI3114 - PCI to 4 Port SATA150):
The BIOS Version of SnapServer was 08.00.13 09/01/06, BIOS of SiI3114 was 5.0.56 dated 10-20-2005. In other Forums I learned, that this BIOS only supports Drives <=250GB, but there were newer BIOS Verisons.
On the Silicon Image Homepage you can find a "abbreviated BIOS intended for integration into system BIOS" 5.4.03 vom 09-22-2006.
With the "HP USB Disk Storage Format Tool" I made a MS-DOS-Boot USB-Stick and with the official AFUDOS.EXE from Ami I then saved the original SnapServer BIOS to file.
With the (probably not so official) MMTOOL.EXE 3.12 one is able to integrate the SiI-BIOS into the SnapServers Ami-BIOS (it's the fourth PCI Option ROM with Vendor ID 0x1095).
Save as, transfer to USB-Stick, flash with AFUDOS.exe and - Tadaahhh !!!

I was extremly happy that this worked for me, so I hope this will be helpfull/usefull to others. (What still is broken is the GuardianOS, but hey, FreeNAS doesn't look so bad after all ?)

[Werner_q]

Originally Posted by blue68f100

Is there a setting in the BIOS to reset to factory setting?

With the null cable setup were you able to watch the complete boot process?

If you have all of the orig hd's. Install them in the order they were removed. Setup a R5 then fail drive 1. May want to do a cold boot in between HD's. Install new let it rebuild once complete do the next and continue through the process till all have been replaced. The delete the R5 reboot, reboot then build a R5.

I have run out of options to try......

I tried exactly as described (but no cold boot between). Seemed to work perfectly, took me about a day with all that rebuilding. When all was fined, I rebooted just to see if it was working ... and nothing. Just the Power-LED lighting steady.
No I know that it was the SATA-controller becoming stuck at boot time.