I'm still waiting for the stats to update. As soon as I can get the exact error message that I searched on, I'll post a link. I looked mainly at the first result which was from Seti. There was a lot of speculation as to what could have caused the problem, but as far as I could see, nothing definitive.

So - a completely different problem, in fact.

Please will you post the messages from your own BOINC log?

Many thanks.

This thread is about abennett's checksum problem. Don't let the title mislead you.

If in doubt, start a new thread.

No need in this case, it's just for future reference.

But we do need more information to solve your problem, because so far you haven't mentioned anything at all. Namely: your operating system, the work unit that had this error, and the message log output. The more information the better, usually.

That was what I was trying to explain:

If the unit is malformed on the server, the checksum would be calculated incorrectly, and no error would show.

If the unit were damaged in transit, then it would fail the download checksum. World Community Grid uses HTTPS, not HTTP. Transit errors are rare, and should be caught at the point of failure (and retried).

So, it remains that the file was corrupted between download and when it was needed. The primary cause for this is file system error.

Since yours is the only report of a checksum error I have seen, this last possibility is a near-certainty.