Yeah. Much better, but not quite there yet. If you recall, I worked at RBTI until I took this position, so I actually worked on the Oterro 3.0 project with the developers to get it as good as it currently is.
The problem up till now was recreating the issue in a way that it was easy to troubleshoot. I think that we have stumbled upon the smoking gun, but putting together a reliable example is difficult because of the inherent randomness of the critter. Mike -----Original Message----- From: Alan Wolfe [mailto:[EMAIL PROTECTED] Sent: Monday, March 08, 2004 2:18 PM To: [EMAIL PROTECTED] Subject: Re: Witango-Talk: Disconnecting Data Sources Have you tried the oterro 3 beta? not sure if it addresses this issue or not though... ----- Original Message ----- From: "Willochell, Mike" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, March 08, 2004 11:09 AM Subject: RE: Witango-Talk: Disconnecting Data Sources Thanks for the suggestions. I don't think that they will do what we need because I have narrowed our problems down to a very weird event. Here's what happens. We are using Tango in conjunction with R:BASE databases with the Oterro ODBC driver. Oterro has known issues regarding its multi-threading model, and it gets scrambled if you hit the data source to fast (refreshing) or if people hit it simultaneously. When this happens, the memory address that Tango is expecting to receive never returns and it sits in an infinite slumber. As far as Tango is considered, the data source is still active, and therefore will never drop the connection. We then need to restart the server to restore functionality. The problem that this causes is that we are imprisoned to monitoring the server to keep things alive. We do not want to create a scheduled task to restart it because unnecessary restarts will cause many database errors and unhappy users. It can go for days without any problems, and then need restarted a dozen times in a single day. I need a way to check to see when the system is hung to restart it. I have explored monitoring the Tango service, but because Tango itself isn't hung, the service does not return anything suspicious. I was going to use a crontab file to do a small database query every fifteen minutes or so, and see how long it took. If the time was excessive, I would be able to conclude that Oterro was hung, and could restart Tango, or better yet, reset all the data source connections so that Oterro could reset itself. That would be the ultimate work-around for now. Mike -----Original Message----- From: Alan Wolfe [mailto:[EMAIL PROTECTED] Sent: Monday, March 08, 2004 1:19 PM To: [EMAIL PROTECTED] Subject: Re: Witango-Talk: Disconnecting Data Sources I'm not sure if this works but you could give it a try. The dataSourceLife setting specifies how long each datasource should last before timing out in minutes. Maybe you can try setting system$dataSourceLife to 0 (which will make all datasources close right after use the docs say) which should close all datasources, then set it back to whatever it was before (default is 30). ________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf ________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf ________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
