LucL
Booting SCSI drive finally expired itself.
The IBM utility had the grace to proclaim it being *faulty*:~/
The myriad of other mounted drives contain my old W98 image
and a separate SuSE8 iteration. Both of these are now working
in the same kit - minus its previously booting SCSI drive of course.
Looks like I have to get my hands dirty...
[update]
Day somewhat occupied with an aunt's 80th birthday celebration.
Tonight and tomorrow we have been told to expect a nasty storm.
Now battening down the hatches here:-)
Dead drive is not really dead but [certainly] has difficulty booting.
Using the above SuSE8 iteration I was able to read various bits of the *dead* drive but was not entirely successful in the attempt to copy/paste its contents to another drive in the nest. However I did drag off some otherwise unbacked up data files which had been orphaned before the appropriate automatic backup had [successfully] taken place:-)
Dug out a venerably old, trusty, noisy, hot-running, Quantum Viking II 7k2 9Gb brick from my goodie box and shoe-horned it into the hole that the absent LVD 10k 9Gb IBM had left. Geoff's earlier concern about my lights dimming when I turned this kit on perhaps need checking out now with this brick running too - four sdds and four hdds a SCSI CDRW and a SCSI MO - all on a generic 300W PSU:~/
Kit now returned to full complement of 8 hard drives.
Unhooked the power harnesses from 6 drives , leaving just two SCSI drives for the somewhat scattergun installation routine of SME server v5.5 to install itself blindly. Afterwards I normalised the power harnesses.
SME is now up in its most basic virgin state. I need to redo all the fstab mounting commands, I-Bays, hdparms and other tweaks. That'll be for tomorrow, while the elements do their level best to try and blast us off the Cornish landscape:~/ No doubt the UPS will get a lot of hits presently...
I very much fear that you, too, are facing an imminent hdd/sdd episode. As I have indicated earlier, even the manufacturers' very own utilities on their own brand of drives, with SMART enabled where appropriate, then they still are virtually useless... IMHO. Trust your ears!!! And your gut feelings too. New drives are JUST as likely to fail as old ones, believe me. There's no such thing as a safe drive:-(
I cannot vouch for RAID, of any flavour, as I have never been tempted to inflict such added complexity on my sanity or to my site. If I could afford a bunch of RAID-able drives I'd much rather use each separately, with diligent backups, and trot out the remainder as and when fatalities might occur! Keeping it as simple as possible is my preferred route. And costs figure rather highly - unfortunately - hence my re-use of a trusty but old Viking while I look around for a present day [equally reliable] replacement.
The intermittent problems are the worst:-| It'd be nice if this sort of thing just happened in a flash, so to speak, the drive fried in a second or two and my ever-faithful nose [literally] could accurately diagnose the source of the problem within minutes:-) Not like I've had for the last few weeks nor, I very much suspect, like you are experiencing. There is another way but it requires a strong nerve. Stress-tests. Find something that invokes the problem reliably [?] and then multiple up the stress until something breaks cover metaphorically and makes a run for it:-) Be ready in waiting with cheque book close by:~/
Night.
best wishes, Robert
16 Replies
1721 Views