Corrupt/Damaged Data Files

[Elmiko]

I run both Empower and Total Chrom, and every so often, maybe once every six weeks, I can't open a previously acquired data file because it's been damaged or corrupted somehow. I don't have any evidence that it happened during acquisition-I just can't reprocess it. A colleague feels this is intolerable, but I've always felt that these things happen when you run under Windows (maybe a Unix environement would be more dependable, who knows), and that one bad file for a thousand runs is no big deal. Am I being naive?

[Mark]

Elmiko,
We have been running Empower for about a year now and have never had a file we could not reintegrate. Occasionally a data file will get "trapped" in the instrument server (LAC/e box) but they can be flushed out to the server easily and are then accessible for reprocessing.

Regards,
Mark

[CE Instruments]

I have no idea about your data system however I think you may have a problem. Datafile corruption is bad and should not be ignored. I doubt it is a software error, more likely a warning that your PC is not well. Back up any critical data and run diagnostics on your hard drive it may be about to die. This type of file error can also occur with badly fragmented hard drives and it may be your file is OK it just cannot be read. Check it out soon before the problem gets worse !

[Tim]

I would concur with CE Instruments. The fact that you mention two different systems suggests it is a hardware (or possibly in part OS related) problem. We had an issue with some of our clients that ran Agilent ChemStation (but fortunately no chromatographic data courruption) which was traced to the hard drives in a number of the clients - they were all the same type and were starting to fail. It was only a particular model of hard-drive too - other PCs bought at the same time and used for the same purpose had different hard drives and did not show the same issues.

This is especially a problem if it is your original data that has become corrupted and even more so if you work in an environment regulated by the FDA (where 21CFR Part 11 applies) - just saying it is OK to acquire the data again will start the bells ringing with them.

[A.Nonymous]

Can you find when it happen?
Like for example when you take an back up over night?

For Empower it is possible that the checksums of the mirror files aren't the same. Then you can lost data if you cant fix the problem.

Maybe you can call to the local Waters Data Specialist?

If it only happens once in 6 weeks, I would be supprised it's the hard disk, then it should be more frequently.

Are you running through a network or is it a stand alone setup?

[unmgvar]

Hy Emilko,

I'll have to seriously agree with what is being said about the regulatory compliance for CFR 21 part 11. If you have to comply to it then you must first of all have all of your data backed up every day. it will anyway solve any data integrity problems you might encounter. better have 2 set of data then one anyway.
if not then you might actually need to see 2 further interesting matter that can also affect data integrity.
the first one is that empower and millenium data files go self destruct if attempted to be opened not through the software itself. check that no one else has access to those file by mistake.
secondly make sur your instrumentand PC are connected to a good UPS and that the electricity in your lab doesn't fall too often. I have seen data being totally corrupted when a power glitch affected part of my lab. the entire data collected went weird. we found out that a faulty connection caused fluctuations in the current and sometimes affected the data.