Hitachi 0F10311 HDS72202ALA330 multiple no spin!

[BrutasMaximas]

I have multiple Hitachi 8x 2TB and a new set of 3x HUS724040 4TB drives after an unknown treatment by a SansDigital box(s).
All the 8x drives spin up just fine mounted in the RAID enclosure and the boot bios of the boxes controller and show 14TB respectively.

If I take the drives out of the enclosure and hook the array to my HBA.
The drives will not spin on the aforementioned model number.
The lets say other 4 spin up just fine.
I'v tried the ones that won't spin were tested individually with and without 3.3V applied.
They just don't spin up. The other 4 in the box spin up just fine.

All of this started for me trying to dump my files from the original array.
I had an original XP drive (yes that's what the software works on) bur could never see the array that I created 8 years ago.
The boot BIOS shows the full array size. The software web page a localhost address show the proper size but any windows XP, 7, 2000 does not.
The files I used to see are not visible.
I really want to recover the files and port them over to TrueNAS, but I can't get past go.
There is no support even after buying 2 systems.

The final point did the box do something to the hard drives that prevents them from spinning up out of their boxes?
The 3x 4TB now exhibit the same behavoiur now that I created an 8TB array on them.

HELP please! Thanks.

 

[sretalla]

after an unknown treatment by a SansDigital box(s).

I don't understand what this is.

mounted in the RAID enclosure

What is that exactly?

the array that I created 8 years ago.

I had an original XP drive (yes that's what the software works on)

Are you saying you used Windows XP to create some kind of software RAID array?

The final point did the box do something to the hard drives that prevents them from spinning up out of their boxes?

I'v tried the ones that won't spin were tested individually with and without 3.3V applied.

So you're already aware of the additional pin trick and tried playing with that?

Generally, if you want some help, you need to do a much better job of describing what we're working with here (hardware and origin story)... it's confusing what we're even starting with from both perspectives. Maybe read the forum rules (link in red at the top of this page) for some hints on how to share your hardware details.

 

[jgreco]

I don't understand what this is.

Sans Digital is a skeevy manufacturer of chintzy grade consumer disk enclosures of all sorts. These were particularly popular with the enthusiast community especially about ten years ago, and back in the FreeNAS .7 days they had a certain popularity in the FreeNAS community before I started heavily pushing on the Supermicro. They were available at places like Fry's and Micro Center, and I believe NewEgg and a bunch of other online retailers. It's probably one of their 8 bay tower units with SAS, since an HBA is mentioned. See

Sans Digital Tower Solution

2.5" Direct Attached Storage Solutions Specifically designed for SSD and 2.5" inch hard drives, these Sans Digital tower solutions offers compact space saving cases in 1-bay, 2-bay, and 8-bay form...

www.sansdigital.com

and I suspect the product being discussed is the predecessor to the

TowerRAID TR8X12G

The TowerRAID TR8X12G is a 8-bay tower with two mini-SAS HD SFF-8644 connections, for high transfer rates of up to 9,600MB/s bandwidth when used with a high performance RAID controller card. Utilizing high bandwidth mini-SAS HD (SFF-8644) cabling, the TR8

www.sansdigital.com

If the user's existing product required XP to manage it, I'm guessing it's pretty old at this point. These units looked pretty nice but getting replacement trays (for example) was difficult, they usually had nonredundant fans, and their power supplies were always gimpy and nonredundant.

Are you saying you used Windows XP to create some kind of software RAID array?

No. Pretty sure it's just a old TowerRAID unit. These were typically just JBOD units that would let you run a pair of SFF-8087 to a RAID controller such as Areca, Highpoint, LSI, etc., cards such as an LSI "-8e" card with the external SAS connectors. Of course this wouldn't have precluded you from doing software RAID instead, and they did have a variety of eSATA, USB, and I think FireWire enclosures. They were very thorough.

 

[BrutasMaximas]

Thanks for responding to my post, it was good to see the first thing in the daylight as I've had a few hours of sleep since my post of desperation.....

The SansDigital TR8M box is an 8 bay, 2TB maximum per tray, software raid yes, with 1ea SiI3726 chips on 2 backplanes equaling a total of 8 bays from 2 eSATA cables to a PC running XP, or 2000. Windows 7 was not yet available.
Here's the rundown of the hardware.

4 ea of eSATA ports connections originate from a RocketRaid 2314 with one cable x 2 to do all 8 trays. The card can handle 2 boxes.
The photo is similar to the one above. The onboard 2314 has one Marvell 88SX7042-BDU1 SATA Port Multiplier IC
(also used in the QNAP TS-419P) to accomplish all the port voodoo.

RocketRaid 2314 card datasheet specs:
• 4 External eSATA II ports at 3Gb/s per port
• Up to 4 SATA II or SATA I Hard Drives (I have Hitachi 2TB SATA III's)
• Support RAID 0, 1, 5, 10, and JBOD
• Compatible with eSATA enclosures
• Native Command Queuing (NCQ) for improved random performance
• Online Capacity Expansion (OCE) and Online RAID Level Migration (ORLM)
• BIOS Booting (INT13) to RAID array

Now for the
The TR8M Specs:
Supports Striped (RAID 0), Mirrored (RAID 1), Mirrored Striped (RAID 10), Parity RAID (RAID 5) modes,
and hot spare on Mirrored (RAID 1) and Parity RAID (RAID 5) mode.
JBOD Software RAID 1, 10, and 5

The TR8M backplanes (2x) have the following one IC per board for a set of 4 drives:
SiI3726 SATA Port Multiplier Data Sheet

http://www.bitsavers.org/components/siliconImage/datasheet/SiI-DS-0121-C1_SiI3726.pdf

Sil3132 SoftRaid 5 Controller
Silicon Image Pseudo Processor Device

And hard drives:

A very recent purchase of new Hitachi HUS724040ALA640 mfg 2015, 3 ea, to do a complete RAID 5 and get the feel of things again since my last go around with the boxes was some 11 years ago now (really?)

The drives were given a full NTFS format to test, if in fact new, was in fact new. They passed perfectly.
The formatting was done on an equally new purchase of a not so new LSI 9300-16 IT mode card. The card is working very well. I was using it for the past week to dump ancient drives onto new 18TB drives and discard the hard drive clutter once an for all. Some had in excess of 20K hours on them, backups of backups etc.

The 4TB drives were moved from port to port to confirm both the card and drive fitness.
Each time the drives were hot swapped with instant startup and various data written and read.
At no point did any of the drives experience the loss of spinup.

On to the TR8M....
I inserted the drives in 3 trays, the minimum for RAID 5, and some 25 hours later a new RAID 5 array was born.
The size however, was incorrect in XP, as it showed some 2TB of space only.
More on that fiasco later.
After much so more, I decided to go back to the original plan, get the data off the 8 drives ( I really have 3 sets, 24 in total).

My goal is to use the original 2 boxes I have, as they were, with the exception of the 4 backplanes, and go straight in with the 9300-16, 2x 4 port cables, as well as daisy chaining the power from the original SMPS in each box. Not a hotswap solution but it's just fine.

I bought 3 daisy chain Molex to 5x SATA socket power cables, and proceeded to do 4 sockets per the 4 drives into the 2 existing Molex connectors in the box from the SMPS.
Using a RAID reconstruction program, no drives were discovered. The drives never spun up. The drives that had been running before the TR8M RAID array was created, now they were silent. All connections were checked with my Fluke. No errors.

Somehow the box has changed the drives. To insult my intelligence further, we have the resulting post topic as seen above.
I took the 8 2TB drives, and did the same thing as the 4TB drives.

In the software, only 4 drives were discovered. The rest were dead cold and motionless.
Back to the other previously functioning box to replug the 8 trays back in and everything is showing up as it should and all drives spin. This not a manifestation of a malfunctioning power supply. All voltages were nominal with 1 or 8 drives attached. The 4 other Hitachi drives of a different model are not affected. As of the writing, the XP machine is unable to see the box through its web interface, The BIOS at boot says the array is 14TB. I am reluctant to pull one of the other Hitachi drives for a model number just right now.

I am lost and confused as to why the 3 and 8 drives don't work without the signal to do so when they did previously.

I can provide photos as needed.

Thank you for your replies......

 

[BrutasMaximas]

I neglected to mention that the 8 LEDs on the front panel as well as the power LED do not light until the BIOS boot has started and then they all can come on with the first of each of 4 first, the the others follow.
Furthermore a staggered start can occur if one wishes, with a varying start procedure. The BIOS confirms the drives are present and their health status.
I think that the hardware or software for the TR8M has modified the drives in such a manor as to force them to be silent until the command has been given to spin. So much for getting my data off with the 9300-16 unless the drives are given the command, whatever that is, to spin.

So I guess that the question is, what could do to force the drives to spinup and be read?

Thanks again.

 

[BrutasMaximas]

Well I guess I don't have to pull a drive to see the other model number as it's in the web page screen shot.
HDS723020BLA642. These drives seem to be not affected to spin up when connected to a power source.

 

[jgreco]

Thanks for responding to my post,

I'm not sure how much help can be offered here. I'm probably one of the better historical resources here as I had at one time considered SansDigital for various potential applications. I'm going to switch to random comment mode in the hopes that more information might prompt insights from someone or help the OP. Do not take my comments as more than "information dump".

The SansDigital TR8M box is an 8 bay, 2TB maximum per tray

I remember that particular box. Came bundled with the RR622 but the 2314 was also available.

The onboard 2314 has one Marvell 88SX7042-BDU1 SATA Port Multiplier IC

This is what made it garbage for ZFS. The nice thing about the RocketRAID cards were that some of them were designed for SATA PM's, which made them a much more attractive option for mid-level numbers of external disks than other options that were based on SAS expanders or other gimpier stuff. SansDigital had designed these things to work with the RR cards, and so they worked pretty well. At least for Windows or bulk storage. However, ZFS does not work well with port multipliers because of its bad habit of sending crushing amounts of I/O to multiple drives simultaneously, which is fundamentally incompatible with a PM.

Sil3132 SoftRaid 5 Controller

Not sure what this is describing.

A very recent purchase of new Hitachi HUS724040ALA640 mfg 2015, 3 ea, to do a complete RAID 5 and get the feel of things again since my last go around with the boxes was some 11 years ago now (really?)

Wow, 50K hours (5.7 years) is commonly considered elderly for drives. I don't know what "very recent" means here though.

not so new LSI 9300-16 IT mode card.

At no point did any of the drives experience the loss of spinup.

Okay. The difference between the LSI HBA and the RR RAID here is that the RR RAID is oriented towards SATA devices, and given the weedy 250W(??) PSU on the SansDigital towers, staggered spinup is almost certainly configured, and if it isn't, it's probably killing your PSU. The LSI HBA can, I believe, be configured for staggered spinup of SATA devices, but I don't think I've ever even looked on the 12G controllers to see.

Staggered spinup can be accomplished by PUIS (drive programming, "Power Up In Standby") or pin 11 (cable control on the power connector). I don't know what method the SansDigital/RR combo uses.

My goal is to use the original 2 boxes I have, as they were, with the exception of the 4 backplanes, and go straight in with the 9300-16, 2x 4 port cables, as well as daisy chaining the power from the original SMPS in each box. Not a hotswap solution but it's just fine.

I really, really, REALLY hope that this means you realize the existing backplanes are unusable trash. The LSI HBA's definitely do NOT support SATA port multipliers. I am not aware of any reason that you wouldn't be able to drag 8 lanes worth of SAS into a gaping hole previously occupied by a backplane port and cable these directly. Be VERY aware that you MUST follow SATA, not SAS, cable length rules if you do this. That means no more than 1M total cable from the connector on the LSI HBA to the plug on the drive.

Using a RAID reconstruction program, no drives were discovered. The drives never spun up. The drives that had been running before the TR8M RAID array was created, now they were silent. All connections were checked with my Fluke. No errors.

I believe that drives that power up in standby would spin if traffic for them were to appear, so this is why I'm concerned that maybe you're not connecting the drives appropriately. However, since you've also replaced the power cabling, this is a horrible mess of possibilities and I am not entirely certain we've established that the drives would spin up when connected to the SansDigital PSU using your new Molex/SATAPower cabling.

I am lost and confused as to why the 3 and 8 drives don't work without the signal to do so when they did previously.

I know you think there's equivalence of some sort. I'm not clear on exactly what is going on either, but it is pretty likely some underlying assumption is incorrect. Been there, done that.

 

[BrutasMaximas]

Well, in response, the support for this is atrocious. I even called SansDigital, and at first, the engineer was helpful but it soon devolved into "We don't support that anymore. We don't have software for Windows 7, but I'll look" It finally resulted in I can't help you.

One answer was of course found on the internet archive. A multitude of options as far as software and user manuals except Windows 7.
Contacting China on the RR2314 card resulted in try this or this... Both rendered the XP to blue screen, with the last .exe resulting in total trashing of the OS. In a total rebuild this time on a SATA drive instead of the original IDE, then reloading from the original CD not the web, resulted in the photos shown above. The CPU is a P4 at some 3GHz or so.

I then built a drive for the same machine with Windows 7 Pro, then installed what was given to me from China. The webpage GUI never contacts the raid array. It sits and spins.

I had the foresight to image both versions of drives before installing the RR2314 and TR8M software to make things easier to nuke and pave.

Regarding the backplanes, I do so intend to chuck them in the bin and go without them, as they are trash to start.
I can't do so yet until there is a resolution to my issue of spinup, access with the aforementioned card and box, and related softwares.....or lack of.
There is one box that has one bad backplane with extended operation time. The BGA controller suffers from the same issues as the heat and serve BGA's I used to do for others at my job with their video game consoles back in the day. 6 months or a year later, they would return and I'd repeat the performance.

A lot of the data and software I get on these 2 products comes at a price of weeks going by searching websites that are no more, or have been scrubbed by their new owners not wanting anything to do with the former companies.

I'm tired and old and don't understand why this is so difficult. The companies that created this in the first place take no ownership. For all the monies spent back in the day by me and others, you would think that they could keep an archive. "Just buy our new whizbang XYZ" I'm having none of it.
That's just one reason why I going TrueNAS where things are supported by you good folk.

But the elephant in the room is the no spinup to get my drives so I can retrieve my files via the 9300.

I did have a working Linux box, 4 core Dell, but I need to buy some 12 capacitors to bring it back. The cost of shipping precludes me to do so until I have a sufficient order. No fake Amazon caps. Only real Nichicon are accepted.
I'm sure that there are a number of things that could be done far easier in Linux.

Thanks for reading and responding,
I'm sure my post will at least provide a good sleep aid.

Thanks.

 

[BrutasMaximas]

"Wow, 50K hours (5.7 years) is commonly considered elderly for drives. I don't know what "very recent" means here though."
No, I neglected to say that the 2 boxes including all 3 sets have about a year on them at most.
The job, always on call, life, moving house and occasional RAID errors relegated them to a corner silent, neglected, dejected and alone in the dark.
Perhaps this is my punishment for their years of suffering.

 

[BrutasMaximas]

Well thanks and well no... Now I feel much worse. So complicated is this business......

Ugh.....
https://www.hdat2.com/hdat2_faq.html#q19

Q19: Hard disk does not spin up - PUIS (Power Up In Standby). ​

A19:The optional Power-Up In Standby (PUIS) feature set allows devices to be powered-up into the Standby power management state to minimize current at power-up and to allow the host to sequence the spin-up of devices. This optional feature set may be enabled or disabled via the SET FEATURES command or may be enabled by use of a jumper, or both. When enabled by a jumper, this feature set shall not be disabled via the SET FEATURES command. Once this feature is enabled in a device, the device shall not disable the feature as a result of processing a power-on reset, a hardware reset, or a software reset. If the device implements this SET FEATURES subcommand and power-up into Standby is enabled, the device shall remain in Standby until the SET FEATURES subcommand is received. If the device does not implement the SET FEATURES subcommand to spin-up the device after power-up and PUIS is enabled, the device shall spin-up upon receipt of the first command that requires the device to access the media, except the IDENTIFY DEVICE command or the IDENTIFY PACKET DEVICE command. Solution for HDAT2 program: if you have a hard disk with enabled PUIS (cannot spin up and BIOS cannot recognize this drive) run program with parameter /W Wake/Spin-up the drive: HDAT2 /W Without the parameter /W, when PUIS is detected on the disk, the program asks you to turn off PUIS. List of parameters: HDAT2 /? or HDAT2 /h.​

 

[jgreco]

Perhaps this is my punishment for their years of suffering.

I have disks that have run for 25 years. Do you really want to talk about suffering?

Well thanks and well no... Now I feel much worse. So complicated is this business......

It is. In the worst cases, I just sit here and try to provide context and random information. There is a good reason we try to guide folks towards a relatively small set of standard selections...

 

[BrutasMaximas]

I'm currently uninstalling all RR & TR8M and their tentacles.
I'll have a go at it in a bit. I will install only from the CD and the 62 page document.

Old like RLL & MFM? I have some of those!

 

[jgreco]

Well I have a nice Fujitsu M2372K 8" HDD up on the wall in my office. I have a display I refer to as the "Evolution of Storage". and I'm happy to have a Storage Module Drive on it, though it might have been cool if I had been able to cough up a 14" one.

 

[BrutasMaximas]

Well those drives beat my collection. Nice.

Now back to the tale of woe and despair.....

I think I know what did in my hard drives to shutdown mode.
It was the files sent from HighPoint in China. The system had been working up to that "update".

Now, of the 8 drives, 4 are showing not active in the BIOS & the webpage, why....because they were turned off in software. No 3V3 pin tied high or any other rubbish that now exists because of someone rejiggering the specs of what a SATA power cable means today, just the lack of the application of an external joomper to shut that green nonsense off.

I feel better now as I try to make a DOS boot disk and apply the Hitachi ftool and reset the bits that started it all in the first place.

While I'm in there, is a drive that -I- someone did that made a 2TB drive into a 1TB drive from ftool way long ago.
That person shall be unnamed because -I- he forgot that -I- he ever did it. I think I'll have a go at it by turning it back to its original size, as it's right in the pdf manual of how -I- he did so.

I also have to ask if you know what the two .05" or .127mm horizontal jumpers are on the back of the Hitachi SATA 1 & 2 TB drives do as I cannot find a reference. All I find is a long manual including hex commands but no reference to the pins even shown in their diagrams in the manual!

Until next post where hopefully this this can conclude with major positive results.

Thanks for the ultimate clue on PUIS. I've never heard of it, let's see if I can make it stop interfering.

 

[jgreco]

I also have to ask if you know what the two .05" or .127mm horizontal jumpers are on the back of the Hitachi SATA 1 & 2 TB drives do as I cannot find a reference.

Sorry, no. Hitachi (also known as HGST) inherited a lot of drive models from IBM back in the day, and with shuffling and reorganizations it is very difficult to locate some types of information these days. The information is almost certainly out there somewhere.

 

[BrutasMaximas]

Even 4TB drives have them except my 3x new 4TB drives where there are 2 empty holes in the molded bits.

 

[BrutasMaximas]

Here I am, I tested a RAID 5 consisting of 3x 4TB drives that can't be seen by the HighPoint card, as each drive as anything over 2TB so it partitioned each drive as so. I learned this after.......

I tried ftool, that only worked to fix the 2TB drive showing as one. It did not fix the PUIS issue, For that I used HDAT and even that needed to be run in DOS, with the HDAT /w to wake up each and every drive.

Then I first tried the RAID 5 consisting of 3x 4TB drives that did not start up. They worked fine in the TR8M box.
Later they were tested by the LSI 9300 card, I used software to recover some files I put there on the RAID box.
Everything worked brilliantly, no file errors. Time to remove the partitioning and NTFS format on all 3. I'm done with you lot. Off to the spares shelf.

Next, I put the 8x 2TB drives in the TR8M box and had errors as seen below. So much for easy things.
Right now it's stuck at 0% and climbing to 0.001%. This appears to be going to take a fortnight.

Right now I have 2 previous defective Hitachi drives being surface scanned-scraped for anything momentous , with one at 90% and the other at 64% and hopefully ending soon.

Be glad you never pulled the trigger on one of these terrible boxes let alone 2 like I did.