Waters LAC/e box Hard drive Failures

[beyondtheprize]

Has anyone seen instances when a hard drive fails on a LAC/e. We are getting this error message:

Windows coudl not start because the following file is missing or corrupt:
<Windows root>\system32\hal.dll.
Please re-install a copy of the above file.

We have successfully resolved the issue by unplugging the hard drive affected and replacing it and using the RAID to mirror the content. My concern is this is our second instance. What can lead to this?

[DR]

First guess: root kit virus
Root kit viruses are easily contracted by those who are allowed to use the internet on LAC/Es and are not careful enough about where they visit. LAC/Es present a particular problem to the IT administrator because if you put too much security software on them, they quit doing their job. Does anyone use your LAC/Es for anything other than Empower related activities? Are your LAC/Es on a netowrk segment that is shared by other computers that may encounter viruses?

Second guess: too much heat, over time, killing hard drive
Newer LAC/Es with dual hard drives and E8400 or similar processors tend to run hot and they're small. It is important to make sure that all vent holes remain well clear of anything that can block air flow and that all vent holes are kept fairly clean. Leaving the left side of a LAC/E up against an instrument is a very bad idea.

Third guess: LAC/E is an older unit that missed the power supply exchange program several years back. Systium (OEM for LAC/Es) has had some serious problems with their PSU vendors over the years. Model 10 LAC/Es (beige) had PSUs that would overheat and kill HDDs replaced with ones that would fail without all the heat, these were later replaced with moderately reliable PSUs. Model 11-13 LAC/Es (black ones) tend not to have PSU problems, but that does not rule them out completely as a cause. Does the skin of the LAC/E w/ HDD problems feel any warmer around the PSU than the corresponding spots of reliable LAC/Es?

Note that these guesses all assume that the content of the Windows message is useless (which I have generally found to be the case when dealing w/ HDD faults).

[Johnny Rod]

http://support.microsoft.com/kb/314477
It's a software not a hardware problem I think, though as DR says, these messages are often little more use than "An error has occurred". Two failures is a bit odd, who installed the OS etc. on the LACE? I would think about doing a clean installation and imaging the LACE. If it packs up it's easy to put back on, and you could use the same image on the other LACEs if they are the same type and it turns out to be a good install.

[benW]

The LACE mentioned is running raid and is therefore should be configuration 12 or later - as such will not be subject to the PSU issue of the earlier configuration LACE.

Hal.dll errors can be symptoms of drive failure. Do you routinely monitoring the health of the RAID array such as via Event viewer? You should be able to see IAANT application event messages with ID 4 that describe the state of the RAID array and these may still be present for you to review.

Two failures from how many LACE’s? As generally hard drive failures are not that uncommon (hence why raid was added to the LACE) even from something as simple as a poorly connected lead. But as mentioned with these small form factor LACE's– especially so with the two hard drives, heat ventilation should be optimised and care to ensure that all fans are working without weird noises and none of the ventilation holes are obstructed. Out of interest, was it the same drive in both cases?

Agreed that this also could be a software failure. Some additional thoughts… Did this occur following a scheduled reboot? If so how often and how are you rebooting your acquisition servers? (some have been known to pull the power out!) XP is susceptible to the boot problems if the LACE was not shut down gracefully.

If this occurred after your LACE needed to be rebooted due to some sort of lock up or failure, then this could have caused your problem. If so, you should consider rebooting your LACE more regularly – LACE’s are recommended to rebooted monthly.

Final tip... LACE’s really should not be used for any other activity than as an acquisition client for the reasons posted, and also because they will require rebooting more often and are likely to be subject to change without any change control (are LACE’s and systems qualified whilst playing solitare, chatting on MSN and manually installing the latest version of IE?).

[EmpowerGuru]

we have hard drive failures all the time.

you can call waters service to replace it or even better, have them send you a hard drive. you can just plug in the new one and upon restart, it will sync them up.

This happens to us enough that we have about ten spares at any given time.

[2panther]

If you need a newLAC/E card you can get them here.