Hi Ken, I manage quite a big configuration : (> 20 machines , 64 instances/machine) So here what I do get things simpler.
1) I have my App to refresh its cached memory in every 10 minutes (with a WOTimer). 2) AutoRecover is always set to "ON" always for all instances. 3) have added a parameter which when true the application's refuseNewSession will get called. 4) Now I have a table which maintains the list of the hostname+posrt combination and a column which says whether the refuseNewSession to be true or false. 5) So after every 10 minutes the WOTimer will get executed and it will start the refuseNewSession for that host and port. 6) Now to deploy my app I have a automated script which does all the deployment across all the machines and at the end it it will make the flag for refusing new session to true. I this way I do not bother to look into the monitor (unless some thing awefull has happened) . Let me know if this helps. I would really like to know if I can improve on my deployment process. Thanks much, Tanmoy On 2/27/07, Ken Anderson <[EMAIL PROTECTED]> wrote:
WO friends, I have to start working on a plan to deploy possibly hundreds of instances of about 10 different WO applications. I currently find it extremely painful to start/stop my existing test instances using JavaMonitor (too much clicking!). I would love to hear from people how they handle large deployments. Do you use JavaMonitor at all? Are you deploying with some way other than on OS X like SSDD or war (I've heard of these things, but have no idea what they are or how to deploy in this manner) ? Any thoughts would be greatly appreciated. Thanks, Ken
_______________________________________________ Do not post admin requests to the list. They will be ignored. Webobjects-dev mailing list ([email protected]) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com This email sent to [email protected]
