Dell 705N Harddrive upgrade

[K3nt]

Hi everyone!
Now I've been reading this forum and others for a while and have some idea on how the thing (705N) works. However, I'd like if someone could take the time to explain the harddrive upgrade to me step-by-step.
The system came with 4x40GB HDs and I got 4x120GBs to put in. Now the system works with the 40GB drives just fine.
When I put in the new drives the system shows up in the Assist program, I can make some settings. But then something goes wrong.

Should I boot up the system with one of the old drives in place? Is there someway to copy the OS using the snap itself?

Is the OS even stored on the hd in this version of the Snap/Dell appliance? It said something about missing language files when saving the inital setup settings. Should I just perform a small OS update to re-upload the language files?

Thanks guys!

[K3nt]

Wow!

I just fired the system up. Waited a while. Then used PowerVault Assist to make an OS "upgrade" and hey presto! it all works.!!!!

[kcrossley]

Originally Posted by K3nt

Wow!

I just fired the system up. Waited a while. Then used PowerVault Assist to make an OS "upgrade" and hey presto! it all works.!!!!

Congratulations! What type of drives (ATA/100 or 133) and speed (5400 RPM or 7200 RPM) did you install? Have you upgraded the OS or RAM yet? What configuration are you using? (RAID 0, 1, 5)? When you get it setup can you post your data transfer speeds here?

Thanks,
Kelly

[re3dyb0y]

Also, did you just install the 4 new drives?

Or did you replace one at a time? Or leave the primary and replace the others?



Yes, the OS is generally on the HD's

[K3nt]

I got 4 Seagate Ultra ATA/100 drives and I replaced them all at once. It took the system somewhere around 20 minutes to be "up". Then it reported some problems. I then perfromed first a factory reset then "upgraded" the OS using the Dell 3.4.790 update and everything works.

Set the drives up as a RAID5 for "security"..

I won't be using this thing for anything else than file storage, so I might leave the OS4 update well alone.

Anyway, I'll post some speed data once I've tested the thing..

[gawganch]

I have also recently upgraded my PV705. I went from 20Gb drives to 120Gb drives and other then upgrading the OS, it worked great. Problem I do have is that network access is very slow. I have tried changing my network setting through the debug channel to almost every scenario. I even took a crossover cable directly to my desktop and still very slow. I am running v.3.4.790, which was the latest OS on the dell website. Any ideas?

Thanks
Steve

[blue68f100]

Originally Posted by K3nt

I got 4 Seagate Ultra ATA/100 drives and I replaced them all at once. It took the system somewhere around 20 minutes to be "up". Then it reported some problems. I then perfromed first a factory reset then "upgraded" the OS using the Dell 3.4.790 update and everything works.

Set the drives up as a RAID5 for "security"..

I won't be using this thing for anything else than file storage, so I might leave the OS4 update well alone.

Anyway, I'll post some speed data once I've tested the thing..

Can you post your HW and BIOS version?

[blue68f100]

Originally Posted by gawganch

I have also recently upgraded my PV705. I went from 20Gb drives to 120Gb drives and other then upgrading the OS, it worked great. Problem I do have is that network access is very slow. I have tried changing my network setting through the debug channel to almost every scenario. I even took a crossover cable directly to my desktop and still very slow. I am running v.3.4.790, which was the latest OS on the dell website. Any ideas?

Thanks
Steve

Have you upgraded the Ram in the PV/snap?
What drive configuration are you running? RAID5,1,0,JBOD

[gawganch]

I have not upgraded the RAM and I am running a raid 5. I was about to blow away the raid 5 and test just as a JBOD and see if I get different results. i was thinking that maybe with the larger drives it was a larger overhead for the raid controller to write the parity stripe. Looks like there is only 64MB of RAM. Do you know what type of memory these units take?

[gawganch]

Product: 2 - "Laser/Spinnaker: 4 drive IDE" - Platform Bytes "2.2.1"
Server-3.4.790, Build Date: Mar 21 2002 20:16:10

PIC Version 1.0, Capibilities 0x00
Failsafe Bios Exists
Failsafe v2.4.437.FS Tue Aug 15 15:23:18 2000

Bios stamp 2.4.437 Tue Aug 15 15:23:15 2000

Plat Bytes 2.2.1

Model Byte 2

Failsafe: -------------------------------L
Flash failsafe sector not locked(30)

[blue68f100]

You are right about raid requiring more memory. But generally RAID 5 will give the best performace with redundency.

Have you looked at the log? For errors.
If it is running JVM, may want to try killing it to see if the performace pickup. The java machine requires a minimum of 24meg of free memory.

Has your disk light gone off? Snaps allow you to connect when the are doing file maintance in background. This has a big impact on performance. Most of the time it shows up on the disk utility status.

[gawganch]

I blew out the raid and used single disks. The performance was 100% better. I am going to re-create the RAID and see if maybe there was a problem with the striping. I don't think There was a JVM enabled, but I will check again. I am not sure what you mean by the statemetn below. Shoudl I look for something specific?

Has your disk light gone off? Snaps allow you to connect when the are doing file maintance in background. This has a big impact on performance. Most of the time it shows up on the disk utility status.

[blue68f100]

Since your speed picked up, I suspect the unit was doing a raid check and repair.

I think you mentioned that you got an error during the upgrade process. Snaps on restart checks the condition of the drives. If the raid had a error. It does a full check of the raid. Which means the controller has to go to the other 3 drives to check the condition of the one in error. This is called parity checking. Which also makes these units with raid 5 so good. There has been some with certain Promise controllers that have a bug, and can not support large disk (LBA48). On some units the array has to be setup with 3 drives and 1 spare. Checking the array may take 12 + hours. Some have reported as long as 3 days with larger drives.

I always like to having the units in JBOD, when doing up grades. In some cases this is the only way to get large disk to work. Alot of the older snap servers store the OS in flashram. Yours is one of those. On these models it is hard if not impossiable to backout upgrades. The new models that use a std boot bios, can backed out by image file.

[gawganch]

Thanks for the info. The only issue I had was when I upgraded the drives, I could not connect to the web interface becasue it stated it had no language support. Once I flashed th OS, it was good. I re-created the RAID 5 and the problem came back. I am doing a format now and will do a disk check and ckeck the log. If this does not help, I will do a 3 drive raid 5 with a HS as you suggsted and see what ahppens. Here are some entries from the disk log file. I don;t see much, but maybe it makes sense to you..This was from the rebuild earlier> I recreated the RAID 5 after the singel drive test was successful.

I File System : Format complete Disk 60000 2/6/2006 10:47:33 PM
I File System : Opened FDB for device 0x60000 Disk 60000 2/6/2006 10:47:33 PM
I File System : All ACLs for device 0x60000, reset to defaults Disk 60000 2/6/2006 10:47:33 PM
I File System : Successfully initialized empty FDB for device 0x60000 Disk 60000 2/6/2006 10:47:33 PM
I File System : Process formatting device /dev/rraid0 Disk 60000 2/6/2006 10:14:32 PM
I File System : Closed FDB for device 0x10018 Disk 60000 2/6/2006 10:13:01 PM
I File System : Closed FDB for device 0x10010 Disk 60000 2/6/2006 10:13:01 PM
I File System : Closed FDB for device 0x10008 Disk 60000 2/6/2006 10:13:01 PM
I File System : Closed FDB for device 0x10000 Disk 60000 2/6/2006 10:13:01 PM
I File System : Logical set device 60000 configured Disk 60000 2/6/2006 10:13:01 PM
I File System : Device 10018 reconfigured as member of RAID-5 set 60000 Disk 60000 2/6/2006 10:13:01 PM
I File System : Device 10010 reconfigured as member of RAID-5 set 60000 Disk 60000 2/6/2006 10:13:01 PM
I File System : Device 10008 reconfigured as member of RAID-5 set 60000 Disk 60000 2/6/2006 10:13:01 PM
I File System : Device 10000 reconfigured as member of RAID-5 set 60000 Disk 60000 2/6/2006 10:13:01 PM

[gawganch]

I recreated the RAID 5, did a format and now the array seems to be going through a rebuild. I't been 10 hours so far and only at 5%. This is most likely why my performance has been so bad. I will let it finish up and see if there is a difference. At this pace it will take several days.

Thanks for the help!

[K3nt]

Hi ya guys!

The HW and BIOS versions of my 705N is pretty identical to the ones posted here in earlier/other threads. 2.2.1 etc. I found no difference then.

I upgraded the RAM to a 256MB module I found sitting in an old dektop machine. Increased the raid cache size to ~80MB but I am too seeing fairly slow network performance. Typically a file transfer to/from the 705N tops out at around 3-4MB/sec, I should be reaching 7-9MB/sec imho.

Seeing it having only a 100mbps nic a theoretical 11MB/sec could be achieved but I doubt that'll happen..

So if anyone has any clues to the network speed issue, I'm listening.. Increasing the RAID cache further?
Anyone tried how high you can set it so it'll still run stable?

I had a problem previously where I set it to 160MB and the thing just died on me, said it had no room to run the admin webserver/pages. 80MB has worked fine so far but I'd like to use the mem to its maximum.

[gawganch]

Hopefully my issue will be resloved when the RAID finishes rebuilding. I have heard on the PV705 (4100) that the switch port you are connecting to should be set to auto in order for the right negotiation. I have a NetGearFS116 switch. Since it is an unmanaged switch, I do not have that option. You can force the 705 nic port using the debug port however. You are getting much better throughput then I was getting.

[blue68f100]

10 hr and 5% means it's having a problem. You said the drives were only 120gigs. That is below the 48bit requirment, should be in 24 bit/std. I think I would stop the array rebuild. The 4100 has 4 promise ide controllers. Make sure all of the jumpers are set to MASTER. Verify that all cables are fully seated. Then try it again.

It should take 10-14 hrs if it is working corectly.
If no improvement change the set up to a3 disk array + 1 spare.

[gawganch]

Blew away the raid and recreated. waiting for the 4 drive raid 5 to format. Then the system will go into building the array. I originally thought after the RAID 5 formats, that was it , but it then does another rebuild. That is what was taking so long.

[blue68f100]

It may be getting errors, look in the log.
It saw the process was interupted. For data integerty it will do a complete check. The formating in unix is one process, after that the drive is setup for raid structure. It may be faster setting it back to JOBD, then raid. It took my 2000 10 hrs to do a 80 gig raid 1 format. My 2200 took about the same time for 300gig. Ive learned to start the process before I go to bed. Other wise I constantly check on it.

CLOCKS RUNS SLOWER WHEN YOU WATCH THEM.

[gawganch]

Still running and only at 17%. I am going to let it run to completeion just to see what happens. No errors in the log. I did just that. Setup 4 disks, then created the RAID. RAID setup took about 15 minutes, then it kicked off the RAID rebuild. Here is the disk log. I don;t see anything funny.

I File System : Format complete Disk 60000 2/7/2006 5:24:52 PM
I File System : Opened FDB for device 0x60000 Disk 60000 2/7/2006 5:24:52 PM
I File System : All ACLs for device 0x60000, reset to defaults Disk 60000 2/7/2006 5:24:52 PM
I File System : Successfully initialized empty FDB for device 0x60000 Disk 60000 2/7/2006 5:24:52 PM
I File System : Process formatting device /dev/rraid0 Disk 60000 2/7/2006 4:51:24 PM
I File System : Closed FDB for device 0x10018 Disk 60000 2/7/2006 4:50:09 PM
I File System : Closed FDB for device 0x10010 Disk 60000 2/7/2006 4:50:09 PM
I File System : Closed FDB for device 0x10008 Disk 60000 2/7/2006 4:50:09 PM
I File System : Closed FDB for device 0x10000 Disk 60000 2/7/2006 4:50:09 PM
I File System : Logical set device 60000 configured Disk 60000 2/7/2006 4:50:09 PM
I File System : Device 10018 reconfigured as member of RAID-5 set 60000 Disk 60000 2/7/2006 4:50:09 PM
I File System : Device 10010 reconfigured as member of RAID-5 set 60000 Disk 60000 2/7/2006 4:50:09 PM
I File System : Device 10008 reconfigured as member of RAID-5 set 60000 Disk 60000 2/7/2006 4:50:09 PM
I File System : Device 10000 reconfigured as member of RAID-5 set 60000 Disk 60000 2/7/2006 4:50:09 PM
I File System : Format complete Disk 60000 2/7/2006 4:49:33 PM
I File System : Opened FDB for device 0x10018 Disk 60000 2/7/2006 4:49:33 PM
I File System : All ACLs for device 0x10018, reset to defaults Disk 60000 2/7/2006 4:49:33 PM
I File System : Successfully initialized empty FDB for device 0x10018 Disk 60000 2/7/2006 4:49:33 PM
I File System : Process formatting device /dev/ride3a Disk 60000 2/7/2006 4:46:17 PM
I File System : Format complete Disk 60000 2/7/2006 4:46:17 PM
I File System : Opened FDB for device 0x10010 Disk 60000 2/7/2006 4:46:17 PM
I File System : All ACLs for device 0x10010, reset to defaults Disk 60000 2/7/2006 4:46:17 PM
I File System : Successfully initialized empty FDB for device 0x10010 Disk 60000 2/7/2006 4:46:17 PM
I File System : Process formatting device /dev/ride2a Disk 60000 2/7/2006 4:43:44 PM
I File System : Format complete Disk 60000 2/7/2006 4:43:44 PM
I File System : Opened FDB for device 0x10008 Disk 60000 2/7/2006 4:43:44 PM
I File System : All ACLs for device 0x10008, reset to defaults Disk 60000 2/7/2006 4:43:44 PM
I File System : Successfully initialized empty FDB for device 0x10008 Disk 60000 2/7/2006 4:43:44 PM
I File System : Process formatting device /dev/ride1a Disk 60000 2/7/2006 4:41:07 PM
I File System : Format complete Disk 60000 2/7/2006 4:41:07 PM
I File System : Opened FDB for device 0x10000 Disk 60000 2/7/2006 4:41:07 PM
I File System : All ACLs for device 0x10000, reset to defaults Disk 60000 2/7/2006 4:41:07 PM
I File System : Successfully initialized empty FDB for device 0x10000 Disk 60000 2/7/2006 4:41:07 PM
I File System : Process formatting device /dev/ride0a Disk 60000 2/7/2006 4:38:26 PM
I File System : Opened FDB for device 0x60000 Disk 60000 2/7/2006 4:26:09 PM
I File System Check : partition is clean. Disk 60000 2/7/2006 4:26:08 PM
I File System Check : Executing fsck /dev/rraid0 /fix Disk 60000 2/7/2006 4:26:08 PM
I File System : Resynchronization of logical set 60000 stopped by system shutdown Disk 60000 2/7/2006 4:22:42 PM
I File System : Closed FDB for device 0x60000 Disk 60000 2/7/2006 4:22:40 PM

[gawganch]

let it run for almost 5 days and it only reached 48%. I am doing the 3 drive RAID 5 now with a HS. Initial results don't look good. After that I will try a mirror. The format runs pretty quick, it's the Rebuilding backup disk that is crawling.