<div dir="ltr">Hi David,<div><br></div><div>I like to leave the event trace turned on. I have found this to be the case at most sites. Also, most workflow people I have worked with agree with this approach. The overhead is minuscule compared to the information it adds.</div>
<div><br></div><div>regards</div><div>Rick Bakker</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Mar 6, 2014 at 3:29 AM, Edward Diehl <span dir="ltr"><<a href="mailto:edwarddiehl@hotmail.com" target="_blank">edwarddiehl@hotmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div><div dir="ltr"><font style="font-size:16pt" color="#000000" face="Garamond" size="4">It's not whether or not the trace is running, it's how often you delete "old" entries and reorg it. Having said that, we have stopped using it to test for duplicate events and now check for duplicate workflows in a check function.<br>
</font><br><br><font style="font-size:12pt"><strong><font color="#000000"><font style="font-size:16pt"><font face="Tahoma"><font style="font-size:10pt">Ed Diehl</font><br></font></font></font></strong></font><font style="font-size:12pt" size="3"><font face="Tahoma"><span style><em>"Success consists of going from failure to failure without loss of enthusiasm."</em> </span><br>
<br></font></font><br><br><div><hr>To: <a href="mailto:sap-wug@mit.edu" target="_blank">sap-wug@mit.edu</a><br>From: <a href="mailto:davidcooper06@icloud.com" target="_blank">davidcooper06@icloud.com</a><br>Subject: Workflow Event Trace Leave On or Turn Off<br>
Date: Wed, 5 Mar 2014 06:55:50 +0000<div><div class="h5"><br><br><div>Hi All,</div><div><br></div><div>The following is more a discussion item!</div><div><br></div><div>I have read in various texts and heard from several workflow administrators that it is recommended to turn the workflow trace off in production.</div>
<div><br></div><div>Reasons:</div><div>1) The trace adds an overhead to the application and database servers, and</div><div>2) The trace fills up the event table(s) with data that is not needed over time.</div><div><br></div>
<div>My argument for leaving the trace running, is more for diagnostic reasons when problems occur in production. It becomes another source for tracking down what happened. Yes the overhead is a given, but I feel this is justified to capture the diagnostic information. As for the database table being filled, implement a deletion strategy which purges the data from the table after a period of say 3 ,6, 9, or 12 months.</div>
<div><pre style="font-family:'Helvetica Neue',Helvetica,sans-serif;font-size:15px;white-space:pre-wrap;word-wrap:break-word">Kind Regards
David Cooper
Linked-In: <a href="http://www.linkedin.com/pub/david-cooper/47/616/36a" target="_blank">http://www.linkedin.com/pub/david-cooper/47/616/36a</a>
</pre></div><br></div></div>_______________________________________________
SAP-WUG mailing list
<a href="mailto:SAP-WUG@mit.edu" target="_blank">SAP-WUG@mit.edu</a>
<a href="http://mailman.mit.edu/mailman/listinfo/sap-wug" target="_blank">http://mailman.mit.edu/mailman/listinfo/sap-wug</a></div> </div></div>
<br>_______________________________________________<br>
SAP-WUG mailing list<br>
<a href="mailto:SAP-WUG@mit.edu">SAP-WUG@mit.edu</a><br>
<a href="http://mailman.mit.edu/mailman/listinfo/sap-wug" target="_blank">http://mailman.mit.edu/mailman/listinfo/sap-wug</a><br>
<br></blockquote></div><br></div>