A31p SATA main drive *PIC*

[Pokrzept]

For my "mission critical" machines, which currently are three A31p systems, I clone the drives (when the mood strikes me; don't have a real schedule) using Ghost. I prefer doing it this way, rather than imaging, as I can simply swap the drives and be back to where I was when the cloned copy was made. Granted, it's not a system which allows me to have daily or weekly backups but I can live with it. For the important files that I need to be able to refer and/or update no matter where I am or what system I'm using, those are on Dropbox.

edit: I should say that for my main home A31p, it's a little bit of an extra challenge to backup. I'm booting from an Ultrabay-connected 320GB SATA drive in the right bay. The main HDD bay is empty although I'm thinking about putting one back in there some day. The left Ultrabay has a DVD writer. The system is always mounted to a port replicator and in turn, there's a external floppy drive attached using the floppy port (not USB). So, to backup I remove the DVD writer, put another Ultrabay SATA HDD adapter in with a 320GB drive and boot from a Ghost 2003 floppy. The other A31p systems are more "normal" in that I use an Ultrabay PATA HDD adapter to clone from the main drive. Those systems are also on port replicators or docking stations using the same type of floppy.

1st.: Sorry for off-top.
2nd.: The way you use your A31p is quite interesting. Can you tell me anything more about HDD-Caddy performance compared to regular HDD slot ? I did some benchmarking under Linux recently and and I couldn't get anything better than 35 MB/s of average speed with an SATA SSD. Thats pretty bad compared to ~~70MB/s that I can get on HM160HC ATA mounted as an regular drive. Performance of ultrabay slot is limited by hardware controller, is it not ?

[rkawakami]

I split this out from this thread: http://forum.thinkpads.com/viewtopic.php?f=57&t=117658

I have not tried taking any benchmarks on disk transfers as I consider the A31p to be slow in most regards anyway. It would guess and say that a SATA drive in an Ultrabay HDD adapter has to be slower than a PATA drive in the main bay simply because of the SATA-to-PATA conversion needed with the Ultrabay adapter. I still should have a bootable PATA drive that I can run CrystalDiskMark on and then compare it with my current SATA drive (Seagate ST320LT012).

[rkawakami]

This is what I got yesterday when running CrystalDiskMark on my Ultrabay-connected Seagate ST320LT012 SATA drive. After setting up this drive last month, I noticed that it was listed in the "Safely Remove Hardware" utility. I didn't think much about this until I looked into how Windows XP had it configured (Device Manager/Disk Drives/Properties/Policies: Write-caching enabled and Optimized for performance. Thinking that this would affect the disk benchmark, I disabled both settings, waited a few seconds and then commanded Windows to shutdown and reboot. The shutdown appeared to go fine but it never booted . I had exactly the same problem last month with my PATA drive, but without changing anything in Device Manager. It was some time since I had re-booted the system and decided it was time. It was slow to shutdown and the subsequent re-boot never took place. All I got then and yesterday was a blinking cursor after the IBM splash screen.

I had tried a number of things with the PATA drive to get it to boot again; FIXBOOT, FIXMBR, using a Windows XP install disk to run a repair. Nothing got that drive to boot. When mounted on another system all of the files appeared to be there. Scandisk reported some errors and seemed to fix them but still no boot. It was then I thought about using a SATA drive so I cloned my PATA backup drive onto it, copied all of the updated files from the now non-booting PATA drive and all seemed well. For about three weeks anyway. At this point I'm not sure if I'm going to continue using the SATA drive. I think I need to figure out why two different disk drives, mounted in two different ways, failed exactly the same way. I've got a strange feeling that something in the software (i.e. Windows) is amiss and is causing this, after all, both the SATA and PATA drives originated from the same disk image.

I recently ran memtest86+ on that system for a couple of hours and it passed so I believe the memory to be okay. I've had some strange issues with this A31p over the last two years so then again maybe it's the hardware. In any event I will probably not be able to get a disk benchmark on a PATA drive right away. I'm going to clone my last PATA backup onto another drive as I don't want to disturb the backup. I am also thinking about swapping my main A31p with a backup system and seeing how it performs.

edit: Member automobus had posted this FAQ on connecting SATA drives to early 2000's Thinkpads last year: http://forum.thinkpads.com/viewtopic.php?f=31&t=115478