Some 1.12.3 issues noticed.

[Mark]

Since upgrading to 1.12.3 from 1.11.1p1, we have noticed the
following changes (non-root pserver, solaris 2.8,
--with-temp-dir):

1) routinely pserver processes exit before cleaning up (or finish
cleaning up) their respective cvsserv-* directory

2) lots of the following two messages routinely being added in
/var/adm/messages:

"Feb 17 14:19:36 batman cvs[29098]: [ID 666953 daemon.notice] bad
auth protocol start: EOF"

"Feb 17 14:41:56 batman cvs[4954]: [ID 719790 daemon.error]
Attempt to write error message before initialization of protocol
buffer.  Message was:  <MESSAGE>]"

one word of the message is in each entry in the messages file...
<MESSAGE>="[server aborted]: received %s signal broken pipe
signal"
 

have not heard any issues from users, timestamps on messages in 2)
do not correspond to timestamps on dirs in problem in 1).... There
is no pattern to messages in 2), so I don't think it can be
atributed to a cron job or some scehduled automated calls to
cvs....

Mark

__________________________________
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html