Intel ss4200-e dead power supply

[Linflas]

I have had this box up and running for years with 4 Hitachi 2GB drives. The power supply seems to have died and appears to be rather hard to find a replacement. According to Intel "The SS4200-E storage system uses the Ext3 file system in data=ordered journaling mode". Would it be possible to take one of my old workstations and boot it up with a Linux thumb drive and get to the data in the array to copy it off?

Edit: Corrected disk size.

 

[you2]

Yes. Ext3 is very common well supported file system. ext4 is the current upgrade but ext3 is fine. You should be able to mount it fairly easily assuming it wasn't in a raid or something that requires a bit more magic. Btw are these 4GB or 4TB ?

 

[you2]

Looking at the manual it is raid 5 by default which is basically 3+1. If they use mdadm (which i suspect is likely but you could contact their support) then yes you will need to install 3 of the disks and make sure mdadm is installed. If you install mdadm on your boot usb it should be able to mount the drives. Looks like it is 4 1TB drives.
--
More information - you might find this thread useful:
https://communities.intel.com/thread/26862
--
Specifically:
The Intel® Entry Storage System SS4200-E operating system volume is mounted as /dev/hda1 and is a Linux ext2 partition. hda1 is a 256MB DOM (Disk on Memory) module installed in the motherboard IDE port.



The data is in an ext3 journaled file system on an lvm2 logical volume on an md raid (RAID 1 for two disks and RAID 5 for four disks) array mounted as /dev/evms/md0vol1. The array consists of a partition from each of the disks. There is only one partition on each disk that spans the entire disk.



The operating system and data partitions use a software RAID 32K Byte chunk size.


The file system block size on the operating system partition (hda1) is 1024 Bytes.

The file system block size on the data partition (md0vol1) is 4096 Bytes.



Did you check to see if the UUIDs were the same for all disks by using the mdadm –examine /dev/sdX1 (where X is the specific disk you want to see: a, b, c or d)?



When you do the mdadm -examine or –detail, what's the "State : " of the array?
---

Anyway as long as it is standard linux stuff (ext3/lvm/mdadm) then yea the data is recoverable it just might take a bit of effort. Be sure to mount the disks read-only.

 

[Linflas]

Thanks for the detailed info, after checking out all the info I decided to see if I could use a regular PC power supply to bring it up long enough to copy everything off of it. I was able to jury rig it up enough to get all my data off of it. The thing has been up and running 24x7 since 2010 so I got my money's worth out of it.

 

[you2]

Interesting. Curious what model disks that thing contained. That's pretty decent if the disks are the original. If you are into linux ubuntu now supports zfs out of the box. One nice feature of zfs and (i think) brtfs is that they will checksum the data so if corruption occurs it will automagically be detected and an email will be sent.

 

[Linflas]

The disks are Hitachi GST Deskstar HD320 2GB drives 2 manufactured January 2010 and 2 November 2009. I bought the NAS when it was posted in the Hot Deals forum here. Newegg says I bought it in November 2009 for $149.99 and the disks in April and May 2010 for $119.00 each. Other than power outages it has been up and running since I put it together then.

 

[you2]

Thanks !

 

[xyvyx]

Hah.. just stumbled into this post... mine died a couple weeks ago too! I probably got it around the same time, and although I had some cheap WD drives in it, it's been doing fine! I've replaced it over the years w/ 2 Synology boxes, but it's still pretty impressive. I've got some ATX power supplies sitting around, so I'll see if I can cobble together something to bring it back to life temporarily (or maybe transplant it into a new box?)

 

[UnCore]

Mine died last year... so then I pulled out the new-in-box one that I had bought for $50 off ebay 7 years ago ready for that day. I assumed it would be the power supply, but if I remember correctly the problem was NOT the power supply. I transferred the OS drive and larger RAM that I had added.

I can't believe the original 4 Hitachi hard drives still work after all these years. I suppose part of that long life is due to the smart power down when not being accessed. I maintain separate backups. But it is still my shared, reliable, speedy, always-on storage.

 

[kscheibel]

https://www.amazon.com/dp/B0015PSYHA gave my SS4200-E a new lease on life. It has served me incredibly well.

 

[Linflas]

https://www.amazon.com/dp/B0015PSYHA gave my SS4200-E a new lease on life. It has served me incredibly well.

Any mods required or is it just a drop in replacement?

 

[tlman46]

I was interested in your answer to Linflas' question. I have ordered the power supply to replace a dead one. From the online description, it looks like I'll need to splice the drive cables from the old p/s to the new one. The old p/s has a 24-pin mobo connector, the new one is I think just 20. Given the lack of pci-e cards, that should not be a problem. Any other considerations to worry about? (I'm waiting for Newegg to ship the p/s.)

ALSO: I was looking online to replace the exact Delta DPS-250AB-24-B. All i could fine domestically were used units ("tested and guaranteed for 30 days" - pretty cold comfort for a p/s with known issues) at very high prices - US$115-175. I say nope! By on a Hong Kong Inventory site, I put a request out and received a number of replies, and purchased one for $50. I paid upfront and didn't use escrow. Fingers crossed - but communication was good.

 

[Linflas]

I went ahead and bought the one from Amazon. There is an additional 4 pin connector that goes with the 20 pin connector for the mobo. It only has 2 SATA power connectors so you will need 2 adapters to connect the additional 2 drives. Also you will need a CPU power extension cable, the one that came with the PS is stretched taut to reach the connector otherwise.

Edit: One other interesting side note to this. Evidently in Windows 10 SMB 1 has been deprecated with the fall 2017 update which is what this device uses to access shares. It turns out that if you had active shares mapped using it then it would not be deactivated so I was blissfully unaware of this change until I turned this device back on after it's long hiatus and could not access the shares. It can be turned back on again in the Windows settings but there is no way I can find to update the device to use the current SMB standards.

 

[SamirD]

It can be turned back on again in the Windows settings but there is no way I can find to update the device to use the current SMB standards.

I just acquired one of these as my first nas and have really enjoyed learning about the very standardized platform underneath it all. Windows server 2008 r2 as a client, I get bursts of 80MB/sec+ until it settles down after a few minutes to just shy of 50MB/sec--not bad for a box that's a decade old.

Mine came to me fully updated and with 2gb of memory, and with the stock cpu (afaik), and this was about as 'modern' as it could be made. I don't think this will be upgraded in stock forum to the newer smb protocols, but the hardware is capable of running other nas platforms with some work. If I needed anything other than this in stock form, this would be a nice platform to build upon. And it's not too shabby in stock form as well.

 

[Linflas]

I just acquired one of these as my first nas and have really enjoyed learning about the very standardized platform underneath it all. Windows server 2008 r2 as a client, I get bursts of 80MB/sec+ until it settles down after a few minutes to just shy of 50MB/sec--not bad for a box that's a decade old.

Mine came to me fully updated and with 2gb of memory, and with the stock cpu (afaik), and this was about as 'modern' as it could be made. I don't think this will be upgraded in stock forum to the newer smb protocols, but the hardware is capable of running other nas platforms with some work. If I needed anything other than this in stock form, this would be a nice platform to build upon. And it's not too shabby in stock form as well.

There are guides on how to install FreeNAS on these but you need to replace the IDE DOM with one with a larger capacity and I believe you will need to use an older version that does not require 8Gb memory.

 

[SamirD]

I haven't looked into too many of the guides deeply, but it seems there's a lot of setups that can be run without the dom and boot of a usb.

What I'm curious about is if this thing can handle drives >4TB and if the 2nd ethernet port will work with the stock setup.

 

[you2]

Is it worth it these days? I mean i don't know how much these cost but I would think naively they are more hassle than they are worth given their age et all. If you have spare parts lying around you can build a system pretty cheap. Mostly you just need some old ram. A pent. class cpu is pretty cheap as well as a low end mb. Biggest head ache is finding a cheap enclosure. Of course if you don't have any spare parts it might get expensive but then in that case i'm not sure you have the expertise to do stuff like run freenas. Anyway having a bit of ram is always a good thing if you run zfs (i presume freenas is zfs based). Not trying to discourage anyone from having some fun getting one of these to run (esp if they found it in the trash); but i normally focus on reliability when it comes to raid.... (unless it is raid 1). I currently run a linux box with zfs raid set and quite frankly the most expensive portion of the system is the disk (but if you use 2 generation old size disk then i suppose the drives are cheap - esp if you find them in a trash bin with this thingy).

There are guides on how to install FreeNAS on these but you need to replace the IDE DOM with one with a larger capacity and I believe you will need to use an older version that does not require 8Gb memory.

 

[SamirD]

There's a couple of things that are nice about these units, the main one being the simplicity of them. Even if the whole thing fails, they are just linux drives that can be put in any linux system for recovery. All the hardware is pretty much standard aside from the custom motherboard, and everything is very cheap these days. I can hit >80MB/s when hitting it via windows server 2008 r2 (win7 64-bit) as a client, and that's really as fast as I need for my data as the file sizes are maybe 100MB at the largest.

The simplest solution is always sans-raid. Raid brings its own issues, especially before the advent of double parity zfs as a single corruption in a raid1 volume would actually be copied to the other drive. I dubbed this spontaneous corruption 'bit rot' and have seen it more and more as areal densities increase while hardware error rates stay pretty much the same (1x10^14/15/16).

Our 'file server' is actually super dead simple--an old xp thin client with some enterprise class drives hanging off it via usb in individual enclosures with fans. And there's 3 drives on each system, the main drive copying new files to the other 2 every 15 minutes. We swap out the drives by year 3 of their 5 year warranty and keep the old ones as bacukps. Not the fastest or most elegant, but been working for better part of a decade now. The intel nas unit will actually be used for an offsite backup via an ipsec vpn tunnel. And since bandwidth across a vpn tunnel is still at best fast ethernet speeds, this nas is actually perfect, especially when in raid1 mode. Plus, you can hang even more usb drives off of it as backups of the backups.

Anyways, that's why I was curious if a 8 or 10TB drive would work. A 4x 10TB drive setup would be 30TB of data, or even 20TB if it's raid1, but all in one quiet compact little device that just works.

I've done a lot research today on adding a second nic. It will probably take a x1 extension and some creative placement to put a small nic in the space there is and then run a cable of sorts to where the stock 2nd port cutout is, but it is very doable. The question for me is if the stock software will be able to use it without data corruption. My guess is that it should if it recognizes the nic.

 

[Linflas]

Is it worth it these days? I mean i don't know how much these cost but I would think naively they are more hassle than they are worth given their age et all. If you have spare parts lying around you can build a system pretty cheap. Mostly you just need some old ram. A pent. class cpu is pretty cheap as well as a low end mb. Biggest head ache is finding a cheap enclosure. Of course if you don't have any spare parts it might get expensive but then in that case i'm not sure you have the expertise to do stuff like run freenas. Anyway having a bit of ram is always a good thing if you run zfs (i presume freenas is zfs based). Not trying to discourage anyone from having some fun getting one of these to run (esp if they found it in the trash); but i normally focus on reliability when it comes to raid.... (unless it is raid 1). I currently run a linux box with zfs raid set and quite frankly the most expensive portion of the system is the disk (but if you use 2 generation old size disk then i suppose the drives are cheap - esp if you find them in a trash bin with this thingy).

Honestly at this point I have gotten my money's worth plus out of this unit so to me it is not worth messing with it further than I have. I have a QNAP TS-451+ that I loaded with 4 WD Red 8Tb drives I shucked from the My Books when they were on sale a couple years back that is my Plex server.

 

[Linflas]

There's a couple of things that are nice about these units, the main one being the simplicity of them. Even if the whole thing fails, they are just linux drives that can be put in any linux system for recovery. All the hardware is pretty much standard aside from the custom motherboard, and everything is very cheap these days. I can hit >80MB/s when hitting it via windows server 2008 r2 (win7 64-bit) as a client, and that's really as fast as I need for my data as the file sizes are maybe 100MB at the largest.

The simplest solution is always sans-raid. Raid brings its own issues, especially before the advent of double parity zfs as a single corruption in a raid1 volume would actually be copied to the other drive. I dubbed this spontaneous corruption 'bit rot' and have seen it more and more as areal densities increase while hardware error rates stay pretty much the same (1x10^14/15/16).

Our 'file server' is actually super dead simple--an old xp thin client with some enterprise class drives hanging off it via usb in individual enclosures with fans. And there's 3 drives on each system, the main drive copying new files to the other 2 every 15 minutes. We swap out the drives by year 3 of their 5 year warranty and keep the old ones as bacukps. Not the fastest or most elegant, but been working for better part of a decade now. The intel nas unit will actually be used for an offsite backup via an ipsec vpn tunnel. And since bandwidth across a vpn tunnel is still at best fast ethernet speeds, this nas is actually perfect, especially when in raid1 mode. Plus, you can hang even more usb drives off of it as backups of the backups.

Anyways, that's why I was curious if a 8 or 10TB drive would work. A 4x 10TB drive setup would be 30TB of data, or even 20TB if it's raid1, but all in one quiet compact little device that just works.

I've done a lot research today on adding a second nic. It will probably take a x1 extension and some creative placement to put a small nic in the space there is and then run a cable of sorts to where the stock 2nd port cutout is, but it is very doable. The question for me is if the stock software will be able to use it without data corruption. My guess is that it should if it recognizes the nic.

My understanding is that the max size drive that you can use is 4Tb.

 

[SamirD]

My understanding is that the max size drive that you can use is 4Tb.

That's the biggest I've seen as well, but there's nothing really different between the 4TB 4k sector drives and newer 8/10TB 4k sector drives except the number of sectors. If I can get my hands on a pair of cheap drives bigger than 4TB, I'm going to try installing them and see what happens.

 

[SamirD]

Honestly at this point I have gotten my money's worth plus out of this unit so to me it is not worth messing with it further than I have. I have a QNAP TS-451+ that I loaded with 4 WD Red 8Tb drives I shucked from the My Books when they were on sale a couple years back that is my Plex server.

For my use case (nas at the end of an ipsec vpn tunnel), they're perfect since they don't have to be blazing fast. It's amazing how much use there still is for equipment if it's quality and in the right scenario.

 

[SamirD]

And these units may even have some untapped capability that no one really tried to get out of them as they were designed for an optional 2nd nic. Now I don't know if an LAG could be set up or anything, but being able to be accessed from two different physical lans would be a real bonus for my use case. I plan to see if a usb nic is recognized before trying one in the x1 pcie slot.

 

[mindless1]

Rare proprietary PSU, I usually repair. Crack it open, see if the capacitors have vented. If not, give it power disconnected from the system, maybe hook up a single LOW-value HDD if it's possible w/o anything else hooked up.

See if there's a PS_On pin (pulled to ground) needed to wake it up, see where the power stops in the path. If not the caps, the switching transistors are 2nd most likely fault, followed by the rectifier diodes. Something something high voltage danger, assess your competence to do this.

Then again if it's several years old running 24/7, even if the caps aren't the fault, I'd at least replace those on the output side of the transformer while it was open.

 

[SamirD]

Usb nic wasn't recognized in any of the usb ports--used a ax88772 chipset based nic I had around.