Brad Cox wrote:

Could someone please advise on what to do about corrupted data,

We need better tools for diagnosing and correcting this sort of thing. In the mean time, I'd like you to send me (or alternatively support@digicool.com) the corrupted data file and the traceback.

and what might have caused it?

This doesn't happen very often (at all, at least not for us). It usually happens when multiple processes open the file at the same time, which should generally be prevented by a file lock. It could conceavably happen if a process is killed in the process of writing a record. I'd like to see the corrupted file to better diagnose the cause and hopefully prevent it in the future.

Also, where is this "/home/brian/temp" stuff coming from? There's no such directory on my system.

This is an artifact of the way that the Python files were compiled. Error traceback show the location where the files were compiled, which is generally different from the run-time locations. Future binary releases will be made in such a way that relative (to the Zope installation directory) will be stored in the files and shown in tracebacks. Jim -- Jim Fulton mailto:jim@digicool.com Technical Director (888) 344-4332 Python Powered! Digital Creations http://www.digicool.com http://www.python.org Under US Code Title 47, Sec.227(b)(1)(C), Sec.227(a)(2)(B) This email address may not be added to any commercial mail list with out my permission. Violation of my privacy with advertising or SPAM will result in a suit for a MINIMUM of $500 damages/incident, $1500 for repeats.

Brad Cox wrote:

Could someone please advise on what to do about corrupted data, and what might have caused it?

Looking at your file, it looks like the problem occurred during Zope startup. Did you have a problem starting Zope just prior to this? The record that contains product information for the MySQL DA was incompletely written at the end of the file. I wonder if a broken MySQL binary could have caused a memory corruption problem that caused the record to have been partially written or caused Zope to crash while writing the record. I have seen memory errors (e.g. bad pointers) cause file writes to fail without errors on Win32. I wonder if something similar happened here. I'm going to look into getting Zope to try to recognize final records that are partially written and ignoring them. The next generation of the database will do a much better job in cases like this. You can always try to repair an error like this by truncating the file at the position reported in the error message. (I do this in emacs, you could do this in a variety of other ways). This makes sense especially if the problem position is near the end of the file. I'll send you a repaired database file (Data.bbb) in a separate email. Jim -- Jim Fulton mailto:jim@digicool.com Technical Director (888) 344-4332 Python Powered! Digital Creations http://www.digicool.com http://www.python.org Under US Code Title 47, Sec.227(b)(1)(C), Sec.227(a)(2)(B) This email address may not be added to any commercial mail list with out my permission. Violation of my privacy with advertising or SPAM will result in a suit for a MINIMUM of $500 damages/incident, $1500 for repeats.