Could anyone with BOINC code access have a look at this one for me, please?

Had a SETI server outage earlier today. But this problem started much earlier: Validate errors II (warning - long thread).

One machine - only - got itself into a state with SETI - only - where every upload ended with:

Error on file upload: no command
[file_xfer] Permanently failed upload of xxxxx
Giving up on upload of xxxxx: server rejected file

None of my other machines was affected (though other users have reported it), and none of the other projects on the affected machine had any problems with uploads. I restarted BOINC on the affected machine during today's total SETI outage, and it has uploaded/reported/been credited since then, so it was a transient problem.

What troubles me is that while the upload problems were happening, about 34 hours, every single upload was terminated as a permanent error at the first upload attempt - no retries, no backoff, just a straight slam dunk. Is that how it's supposed to work?

BOINC is v5.10.13, Windows 32bit, running on a Windows 2003 server as a service (but I have other W2K3 servers which didn't suffer the problem).

I'm trying to gather further information on this problem in a new SETI thread: Rash of Validate errors

So far, we've had one report of it still happening under BOINC v5.10.30, but the wording has changed:

[error] Error on file upload: no command
Giving up on upload of xxxxx: fatal upload error

---