My previous post declaring a successful resolution to the problems I was experiencing with Parallels 9 VM’s stored on an external USB HDD proved premature. The apparent success of being able to start / suspend / resume / shutdown / restart etc, something that had previously been impossible, was sadly misleading though my conclusion that Parallels 9 was not responsible still appears to be valid.
Upon returning home last night and sitting down to resume my exploration of Android AppWidget development with Oxygene, once again my VM’s were refusing to start with the same old permissions complaint.
What on Earth was going on !!?
Fine. I had some other file housekeeping to attend to, so I set about taking care of that, only to find that I couldn’t even create new folders anywhere on the affected volume (that now all too familiar “error code -50”) !
Trawling through the now also familiar list of google results I found a couple I had not previously visited and came across an entirely new suggestion. Nothing to do with permissions or ACL’s or file systems or anything so obvious.
No. This related an experience with energy saving HDD able to place itself in stand-by/sleep state independently of any OS, and the fact that if OS X Energy Saving settings are not themselves set to suspend HDD’s to save power, then when the OS is confronted with a HDD that is/has been suspended, it seems to get all discombobulated.
And it occured to me, that I use Western Digital “Green” drives. Not through any sense of responsibility for the environment or a crusade to drive down my power bill, just because they are cheaper and quieter and the less-than-bleeding-edge performance isn’t going to make that big a difference in my life. 🙂
I have a hazy recollection of having specifically turned this OFF in the distant past, precisely because of the opposite problem – that the disks being put into standby resulted in them not properly re-awakening.
As I mentioned previously, the physical drive now is different from that before – 2TB rather than 1TB, WD Green in both cases. But the drive was changed some time ago and the problem only started happening this past weekend. I can only think that the OS X 10.8.5 update contained some change to the behaviour in this area.
Update: Although not describing exactly my problem, this seems to confirm that the issue was caused by the OS X update and that a fix is already in the works.
Having turned this Energy Saver setting on, since I rebooted an hour or so ago the problem has yet to recur.
There is a definite depression forming on my desk where I am constantly touching the wood there!
Just my 2 cents, but Western Digital “Green” drives are probably one of the worst out there. Here is a link to one of the sites describing a problem:
http://news.softpedia.com/news/WD-Caviar-Green-HDDs-Suffer-From-a-Critical-Design-Flaw-196159.shtml
I guess you are experiencing just that. Also I have bad experiences dealing with WD drives in general.
Ok, reading again you seem to have a recent enough drive that probably does not have a problem in the link anymore.
Everyone has their own HDD war stories. I can’t say that the problem reported in that link is something I encountered even with the previous 1TB Green drive for whatever reason. 🙂
My worst experience was with Seagate, when I fell victim to their firmware bug where HDD’s would brick themselves if the number of bad sector errors hit a certain number, with no way to recover the drive. Prior to that, a Hitachi drive that just up’ed and died on me. Unfortunately, because it was the boot drive in an Iomega NAS RAID unit, and the RAID was a software RAID implementation, without the boot drive all that lovely redundant data was inaccessible.
WD have for me – touch-wood again – thus far been reliable. Then again, there was a time when I would have said the same thing of Seagate and Hitachi drives. 🙂
I agree, it comes to personal experience. As for RAID and data safety I can only say you have to be very careful. Even HW raid can fail and you then have the same problem. That is why RAID 1 (mirror) is the best for safety. Even if RAID fails (SW or HW) you can just get the data from one of the disks. RAID 5 is another matter. That is why I have a rsync scheduled from my RAID 5 array with my personal data to a RAID 1 pair of disks. That way it is almost impossible to lose data.