Bought new disks, any simple way to migrate the pool?

[Octopuss]

I ran into a stupid annoying problem with 3,5" to 2,5" adapters and have to put the old disks back until I solve this.
I have replaced the HBA already though, and I can't import back the pool, because none can be found. What should I do? When the system was booting, I believe I saw what looked like initializing the existing disks.

 

[Octopuss]

Actually, nevermind. The disks aren't even spinning, but I can't figure out why.
The card works for sure -it's detected and passthrough is enabled. I also had it in my PC when I updated the firmware.

I noticed there is a tiny click when I connect the cables to disks, but they wouldn't spin up. This could be a cable problem, but why wouldn't it work?
It's this cable https://www.ebay.com/itm/234698421266
The HBA is LSI 9305-16i

 

[Octopuss]

Ok, I am getting somewhere.

edit: I should have used sas3flash, then the card would show up.

But on the ESXi's side, the card is passthtrough'ed.

 

[Octopuss]

But also - monologue while googling up more info continues - there's this:

 

[Octopuss]

I have plugged the card back into my PC and tried connecting disks to it with the same result: barely audible click instead of disks spinning up.
I am 99% sure the card is ok, which leaves me with the cable being the culprit. But what could possibly be wrong? 8643 to 8482 is exactly what I need and that's exactly what it is. Buying a different one will most likely result in getting exactly the same thing, because there aren't really any variants of this.

 

[NugentS]

Try it on bare metal rather than as passthrough -see if it works then

 

[Etorix]

The disks aren't even spinning, but I can't figure out why.

How are the drives powered? And how were they powered when they worked?
This could be the "3.3W Power Disable" issue.

 

[Octopuss]

Try it on bare metal rather than as passthrough -see if it works then

Have you seriously not read the post directly above yours? WTF man? You literally replied to me saying I plugged the card into my own PC and tried connecting the disks!

How are the drives powered? And how were they powered when they worked?
This could be the "3.3W Power Disable" issue.

Few posts above I linked the cable.

Pardon Our Interruption...

I remembered I had exactly the same problem with the current setup when I bought SFF-8087 to 8482 cable that had SATA power connectors. That was fixed by buying cable with molex connectors (something like this), but that's what I have right now and the disks (and the SSDs too) still won't power up, they just do a click.
I really have no idea what else to try, I'm out of ideas.

 

[Etorix]

With 4-pin power connectors, it's not the 3.3V issue. I'm out of ideas as well, and I suppose you have already checked and recheched that everything was properly seated.

 

[Octopuss]

Yep, I tried in two different machines with same result.
I wish I could buy these cables locally, because returning a cable bought from Ebay is a major pain in the ass and almost pointless from financial point of view because of shipping cost :(

 

[MisterE2002]

Have you seriously not read the post directly above yours? WTF man? You literally replied to me saying I plugged the card into my own PC and tried connecting the disks!

A bit of respect please. People are trying to help you here. We know it can be frustrating if things do not work.

 

[Octopuss]

Respect? I think not bothering to read even the very last post of a thread is blatant display of disrespect to the OP.
Your argument is invalid.

 

[artstar]

Yep, I tried in two different machines with same result.

Despite what the eBay listing for that cable claims, I'm still inclined to think there's something funky going on with the 3.3V Power Disable feature that your drive likely supports. If you have access to a multimeter, can you check to see if pin 3 of the SATA power connector has any connection to any of the pins of either the Molex or the mini SAS HD connector? Better yet, if you have the luxury of plugging in a drive and measuring the actual voltage on pin 3, that would reveal the vital clue.

If it is open circuit or connected to ground, then it should definitely be spinning up but if there is continuity to anything else that could possibly carry 3.3V (or higher) to it, then that will be your likely culprit.

 

[Octopuss]

The thing is, I put the old HBA back, and the disks work just fine.
As far as I can tell, the only difference between the cables is SFF-8087 connector on the old on and SFF-8643 on the new one. Both use molex connectors for power, so the power disable thing does not apply even if it was present (it's something implemented in the device or power supply, or?...)
I can try some regular 2,5" HDD from an old notebook too, but I think it cannot be related to anything I connect to the cable. It just wouldn't make sense.

And no, unfortunately I don't have a multimeter and don't even know how to use one. Physics is something I avoid like a plague.

 

[artstar]

The thing is, I put the old HBA back, and the disks work just fine.
As far as I can tell, the only difference between the cables is SFF-8087 connector on the old on and SFF-8643 on the new one. Both use molex connectors for power, so the power disable thing does not apply even if it was present (it's something implemented in the device or power supply, or?...)
I can try some regular 2,5" HDD from an old notebook too, but I think it cannot be related to anything I connect to the cable. It just wouldn't make sense.

And no, unfortunately I don't have a multimeter and don't even know how to use one. Physics is something I avoid like a plague.

The 3.3V power disable is implemented on a chassis level so that techs running data centres can hard reset HDDs remotely but I'm suspecting something a little oddball in your case. Yes, one would hope that the cable you purchased does indeed take away the 3.3V from pin 3 of the power connector but that's not a guarantee and the fact that the end connector for the drive is all one piece doesn't give me confidence that pin 3 of the power connector is indeed isolated from the mini SAS HD connector. Hence, to be really sure that it is doing what's expected, a multimeter will confirm the result.

https://documents.westerndigital.com/content/dam/doc-library/en_us/assets/public/western-digital/collateral/tech-brief/tech-brief-western-digital-power-disable-pin.pdf

The fact your old HBA works adds credibility to Etorix's (and my) theory on this.

If you don't want to purchase a multimeter that costs about as much as your last cable did, you can always buy two separate cables - one that has only the mini SAS HD breakouts and another that has only the SATA to Molex power connector.

 

[Octopuss]

Are you saying there is a chance it's the HBA itself that does this?

 

[artstar]

Are you saying there is a chance it's the HBA itself that does this?

I would not expect it to, as the specification typically suggests the chassis is responsible. WD's documentation suggests the same. However, given the circumstances, I'm definitely suspecting some weirdness involving the HBA. That's why separating the connectors entirely might be the likely answer here. In physically isolating the mini SAS HD breakout from the power connector, only then can we be absolutely sure that the Power Disable pin has been truly isolated.

Of course, I like to test my theories with scientific methods (i.e. multimeter) to prove myself right or wrong before going shopping. :)

 

[Octopuss]

But why would the data wires be connected to the power ones? Or what is it you're saying might be happening? And wouldn't that mean the cable is at fault?
I mostly don't have the slightest idea what are you saying to be honest :D I see a cable that splits and clearly see what wires go where. How can that interfere is beyond me.

 

[artstar]

But why would the data wires be connected to the power ones? Or what is it you're saying might be happening? And wouldn't that mean the cable is at fault?
I mostly don't have the slightest idea what are you saying to be honest :D I see a cable that splits and clearly see what wires go where. How can that interfere is beyond me.

What you're not seeing is the connections within the plastic. Is it a PCB that has a trace going to that pin? No way of knowing without either shaving down the connector or checking with a multimeter.

Does that mean the cable is potentially at fault? Absolutely. Hence the ultimate solution being to physically separate the data and power connectors, to give us absolute certainty of that isolation.

Why would data wires be connected to the power one? Well, in this particular case, the Power Disable pin is purely a logic signal. It draws extremely little current (think microamps), as logic signals typically do. Just because it's a power connector doesn't mean that all pins need to be drawing large amounts of power.

The biggest struggle I'm having with this adapter is trying to find a manual, in the event that it's something simpler, as in a setting in its own BIOS. Broadcom are about as useful as an ashtray on a motorcycle when it comes to documentation, not all that different to other major gear I've worked with outside of the computing realm.

 

[artstar]

I don't know how relevant this manual is to your controller, based on my search for Avago SAS3224, but there is an interesting feature mentioned on page 15 called "Dimmer Switch Features". https://www.cisco.com/c/dam/en/us/t...4385-00_RevF_12Gbs_MegaRAID_SAS_SW_UserGd.pdf

It suggests there is a Dimmer mode (Dimmer Switch I) which spins down unconfigured disks. It can be disabled. Page 84 makes mention of this. Maybe try that first?