[imapproxy] nulls...

Eric Rostetter eric.rostetter@physics.utexas.edu
Sat Nov 9 00:35:32 2002


I reported earlier that I was having problems with messages with nulls in them
hanging up my imapproxy/IMP install.  It has somehow now stopped hanging -- see
below for details.  But first, a recap...

It was indeed a problem I *was* seeing.  When it was happening, I was seeing 
this in the debug output:

[DEBUG 19808] parent_locate_and_activate_record(): Free record found for
"ericr". Reactivating ...
[127.0.0.1:3163 (3092)]  Reactivating record for user "ericr" (9).
[DEBUG 19808] Forked a child 20482
[DEBUG 19808] parent_conversate(): select() timed out. (60 seconds without activity)
Status: 0 temporary, 1 active, 1 inactive, 0 other, 2 total.
[DEBUG 19808] parent_handle_idlers(): Sending NOOP to server -1.
[DEBUG 20482] child_conversate(): select() timed out. (60 seconds without activity)
[DEBUG 19808] parent_conversate(): select() timed out. (60 seconds without activity)
Status: 0 temporary, 1 active, 1 inactive, 0 other, 2 total.
[DEBUG 19808] parent_handle_idlers(): Sending NOOP to server -1.
[DEBUG 20482] child_conversate(): select() timed out. (60 seconds without activity)
[DEBUG 19808] parent_conversate(): select() timed out. (60 seconds without activity)

Which repeated on with lots of select failed, etc.  

Anyway, I tried to debug it.  I rebuilt/reinstalled the imapproxy binary
with some additional debugging, and the problems "went away."  Rebuilt again
after removing my debugging, and no hanging.  A "cvs diff" shows no changes.
So maybe I just had a bad build that was causing these hangs with nulls???

Anyway, now the behaviour is still broken, but it no longer hangs up and 
freezes the session like before.  Instead, I get the message up until the
first null.  It doesn't return anything after the null. This still isn't
correct behaviour, but at least it doesn't freeze up any more.  And is
much more expected behaviour at least...

Well, I'll let you know if anything changes, or I find out more...

-- 
Eric Rostetter
The Department of Physics
The University of Texas at Austin

Why get even? Get odd!


More information about the imapproxy mailing list