Topic: I think it's dead:-|

[Robert Harlow]

[update]

Well, I managed to push a small picture file (7Mb) across the network onto the SME box. Never got around to checking it but it DID complete.

Tried pushing across a 113Mb mpeg but it stopped before the end with a network name no longer availabe, so then I deleted it.

Tried pushing across a 200Mb mpeg... it stopped before the end with a network name no longer available. Haven't managed to delete it or do anything else.

From the W2kPro box's Explorer I can't even see the SME box anymore nor can the . There are no network ligthts flickering. I cannot wake the SME's monitor to display anything. SME's keyboard light does respond. And the SME's IDE LED is in free-fall.

I sense a manual enforced hardware reset button episode spawning the usual hard drive errors, corruptions, data risk etc etc. Really stable this SME box.

I think it's time to run the manufacturers' hard drive utility floppy once again on all the drives - so that it can tell me everything is fine just before I hear a dull clanging tizz that REALLY tells me that everything is not fine.

BTW a couple of nights ago I ran MEMTEST86 fully, it went around the loops 3 full times with no errors.

The BIOS reports all voltage levels are up to their usual levels.

The UPS is running as per usual for the network and all the boxes.

Anyone from SME who wants to get involved then just step right in... I won't bite your head off [well, I hope I can control myself adequately] but I will continue to post the ways that SME is both trashing itself, all of my data, perhaps my drives too and most of my time... until it returns to running quietly, reliably, unattended - as it's supposed to do as it has done so before for months on end.

best wishes, Robert

[Robert Harlow]

[update]

Hope I've tracked down -all- the problems.

Some 6hrs of intensive hard drive tests finally spewed out a drive with a problem - the second of a pair of 10k LVD SCSI drives, both have now been decommissioned and destined for the great IBM scrapheap in the sky:-| So, only two drives down... so far:~/ But I haven't heard any more *noises* so that's certainly a plus .

The onset of the trouble appeared to be loosely to be allied to network activity in excess of a brief lookup ie bunches of files and the like. The switch's port and CAT5 cable swapped out OK. The NIC was fully inserted and really clean. This was the most recent NIC I bought and all the contacts are immaculate. Handling it I thought it seemed a bit *thin* (physically). Either that or the PCI slot springs too weak. It fitted but without much effort to seat it... Tried the nearby empty slot  and it needed much more effort to seat similarly. SME fired and stayed up... No more issues with long files or large batches of files or strangely intermittent [read none showing when things should've been running full blast] network activity lights, the network activity is back up solidly.

Have been stress testing this iteration of SME for two hours now without it falling over. Hope I have all the issues identified now:~/ Two intermittent SCSI drives - one booting - the other, rather unfortunately, also used for swap... And a poorly gripping PCI slot running the NIC. Just what the doctor ordered for certified mayhem.

I will now isolate all my remaining drives bar the one SCSI  that'll run SME the best, re-install and hope I can muscle up enough backups to cope with the catastrophic data loss of what was my reliable centralised data repository.

Thanks for listening people and to especially to those who helped me privately behind the scenes. If anyone from Mitel is lurking, while I have your attention, please add two wishlist items to your workload...
* long file handling [read in excess of the dire 2Gb limitation]
* a vastly more intelligent/inter-active installation routine [ditch the scattergun technique]

best wishes, Robert