I had issues with 5.0, on windows, related to witango opening
connections. It seemed there were issues when more than one witango
thread, were opening a connection at the same time. Usually
happening, in the morning, going from a time of little traffic, to a
lot of traffic. I did a BUNCH of testing, and troubleshooting.
NOTHING fixed the problem, except moving to 5.5009.
You can look for a real long message I did, hammering 5.5 with
regards to this issue.
I limped along by playing with the datasourcelife. set it to like
24hours and the crashes were MUCH less frequent, cuz the connections
never shut down. I could also make the crashes a lot worse, but
making dslife 0, cuz it makes witango open and close the connection
on each query, no caching.
Unfortunately, I don't think you are going to get a better answer
than upgrade to 55 and 10.4, or don't use witango. At least on
windows, 5.5 made those problems go away for me, and I tried to work
around it with v5 for MONTHS.
--
Robert Garcia
President - BigHead Technology
VP Application Development - eventpix.com
13653 West Park Dr
Magalia, Ca 95954
ph: 530.645.4040 x222 fax: 530.645.4040
[EMAIL PROTECTED] - [EMAIL PROTECTED]
http://bighead.net/ - http://eventpix.com/
On Dec 1, 2005, at 4:44 PM, Roland Dumas wrote:
On Dec 1, 2005, at 3:24 PM, Phil Wade wrote:
I hope this explains the situation on why we are unable to decode
backtraces for Witango Server 5.0 on OS X more clearly and why
debugging your 5.0 problem is going to be trial and error and may
not be able to be solved without upgrading the OS, ODBC, DB and
Witango Server.
cutting to the chase:
I accept that there isn't a system available to customer support to
trace out a problem in a previous version of the server and that
I'm in a trial and error situation. Also, I know that the problem
might be a bug in one of the systems that would be replaced by
better systems were I to upgrade everything. Upgrading everything
to solve a bug is a non-trivial thing to do in terms of time, cost,
etc., and it might not cure the problem if it's in the class of !
CST corruption errors introduced by the studio.
That's why I'm looking for ideas to try that I haven't tried yet.
For something this odd to be going on with a relatively simple
action is perplexing. It was not triggered by the OS or ODBC
upgrade, but by the simple addition of a field in a table. For a
simple query to nuke a server only when a record is hit the first
time and then only randomly, suggests that something might be funky
in the database and it's passing back something that gives witango
indigestion when mysql is updating its internal index. All that
stuff is beyond me.
Looking for ideas that are possible to try short of shutting down
the business for clean installs during the busiest season of the
year._________________________________________________________________
_______
TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
________________________________________________________________________
TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
