Dropped Hitachi - Live PCB swap - End of the road?

[gvm]

I have a dropped Hitachi HDS721010KLA330 (1Tb) that contains some "useful" data, but not "critical" - a small no of files that weren't backed up at the time. I've been through a few basic DR processes with it, with no success, and I'd like to get opinions on whether there's anything else worth trying before I declare it dead.

What happened was that the drive was running, in a metal Icy Box that was standing vertically (on a plastic stand) on top of a tower case. It fell pretty much vertically backwards, and the corner of the metal case embedded itself vertically in the wooden floor, after dropping a couple of feet. A rough calculation gives a deceleration of about 200G, in the plane of the platters. The disk went offline, then came back about 30 seconds later. It then seemed to work for around 20 mins, just probing around and reading data with dd, at which point the system blue-screened. From that point on, it's been dead.

It now spins up, but with the dreaded regular clicking. MHDD shows it as constantly "BUSY" from the moment it powers up.

I had hoped that the fact that it continued to work after the fall pointed towards a trashed SA but functional heads, and that a live PCB swap might work, so I sourced an identical drive with same firmware, and tried this.

What I've tried is booting from a ddrescue-capable gentoo-based live CD, with the failed drive's PCB connected to the donor disk. This works OK, pointing towards donor/patient compatibility. I suspended the drive using "hdparm -y", and moved the live PCB to the patient disk. I then "woke" the drive by trying to access it (using dd, smartctl and various others). At this point it spins up the patient disk and, unfortunately, goes into click mode.

So, question is "is it possible that it's still trying to reload the SA when it's revived from standby, or is this the end of the road?" As I mentioned, the data is "useful but not critical", so it's not going to justify the cost of a head swap.

Thanks

[dobrevjetser]

Game over.

First of all hitachi's are a nightmare for hotswap, and the way you are doing it is never going to work.
Second, your drive probably had light media damage that got worse while using it in the 20 min after the fall.
You should have copied your data at that time.
Thirth, the media damage killed the heads and now you have a dead hitachi with bad heads AND media damage!
Even a head swap will be very difficult in this case.
Sorry to bring bad news.

[networks]

The head swap is what is going to get it going "if" no platter damage occurred. The head stack and or platters sustained the brunt of the damage on the fall. This isn't a "software" or "controller" board failure. That's just "my" opinion I am sure others will chime in as well.

[gvm]

Thanks for the repies - both dobrevjetser and Spildit have said that I'm doing the live swap wrongly. Could you be more specific?

[mr_spokk]

Why try to hotswap a drive that have felt over, and have broken head(s) and probably a media damage?!?

[dobrevjetser]

Moreover on modern drives hotswap involves a lot more. Dont forget drives have a plist and adaptives which is different for each drive.
Like mr_spokk says this is of no use if heads and / or media is damaged.

[gvm]

Why try to hotswap a drive that have felt over, and have broken head(s) and probably a media damage?!?

Because the drive ran successfully for 20 minutes after the impact - making broken heads relatively unlikely. Obviously, they could have broken at some later point, but it was worth the attempt, just in case the impact had just gouged the SA.

[gvm]

Yes, read my last post.

Plug the donnor with the native donnor pcb
Spin it down, and without power off move the donnor PCB to the damaged drive.
then spin it up. It's the donnor drive with donnor PCB that have to be initialized, then put to sleep. Then you move that PCB to a damaged drive and hopefully will "awake" without the need to read SA because the SA was initialized already.

The way you are doing you are using a non-native PCB without NVRAM transference on a drive and that is not going to work.

Thanks for the explanation - apologies for skipping your post, I guess we were both editing at the same time.

I tried the swap using the donor PCB instead of the native PCB, and it had just the same result.

[Arthur]

Because the drive ran successfully for 20 minutes after the impact - making broken heads relatively unlikely.

On the contrary, it is very likely. Seeing cases like this almost every day. Don't make any further attempts unless you want to make it more worse than it already is.

[pcimage]

Hot swap on this drive will NEVER work, even if the heads were good.

Drive has physical issues, so end of the road for DIY

[mr_spokk]

Why try to hotswap a drive that have felt over, and have broken head(s) and probably a media damage?!?

Because the drive ran successfully for 20 minutes after the impact - making broken heads relatively unlikely. Obviously, they could have broken at some later point, but it was worth the attempt, just in case the impact had just gouged the SA.

Sometimes when you run over a sharp object with your tire it goes flat instantly, and sometimes it takes a few miles...but it will go flat.
The same happens with your drive, you get a little damage on the surface on impact, and after 20 minutes it has ruined your head(s).

It's the end of DIY, that drive need a new set of heads and cloned professionally...sorry to say.