something you might try too is a variation of a technique we have used in our system to better track crashes.  we dont have a crashing problem but we do have a crash every now and then.

what we did was make a table in the database called "Benchmarks", which had an autoincrementing ID, a text description, a start timestamp and a stop timestamp.

At the top of every taf file, insert a row into benchmarks, using the name of the taf file (<@appfile>)  for the description field and <@currenttimestamp> for the start timestamp.

Store the ID generated in a local var

At the bottom of each taf file, update the row in benchmarks by id, setting the stop timestamp to <@currenttimestamp>.

as is probably obvious, what this will do is let you know which taf files are crashing and which arent (well atleast you will be able to see requests that don't finish) by identifying rows which have a blank stop timestamp.

If there is one consistant .taf file that continualy crashes you will know where your problem may lay.

If the problem isn't apparent, you can log more things in the benchmark table to help your investigation, such as vars and their values.

using <@varnames> and looping through the results for each scope, storing the names of all variables and their values in a large text field (as well as <@argnames>) might give you alot of clues in what's going on and causing the crash.

Even if it is witango or your database software crashing, this should atleast be able to help you help either software company to help find and resolve the issue, or atleast give you a workaround.

When we did this we found right away a couple problem peices of code, and what else is neat is we added a column that computes stoptime-starttime so we can see what requests take longest and optomize them as needed.

Hope this helps!


On 10/3/05, Robert Garcia <[EMAIL PROTECTED]> wrote:
Mark Weiss was having a ton of issues, over and over. Ask him about
what seems to be working for him.

--

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 Oct 3, 2005, at 6:26 PM, Wayne Irvine wrote:

> Here's a crash log entry covering one crash.
>
> I've looked through it and the pertinent bit seems to be the last
> line:
>
> Link (dyld) error:
>
> dyld: /Applications/WitangoServer/5.5/./witangod can't open library:
> libjs32.dylib  (No such file or directory, errno = 2)
>
> Anyone know what this is?
>
> Wayne Irvine
>
>
>                   Byte Services Pty Ltd
>                http://www.byteserve.com.au/
>                   [EMAIL PROTECTED]
>    Ph 02 9960 6099   Mob 0409 960 609   Fax 02 9960 6088
>
>
> ______________________________________________________________________
> __
> TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
> <crashlog.txt>

________________________________________________________________________
TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf

________________________________________________________________________
TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf

Reply via email to