If you can suffer the consequences and it is only ONE job, stick it in SYSSTC. Your mileage may vary on this one................ But monitor EVERYTHING else closely...
Claude -----Original Message----- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Vernooy, C.P. - SPLXM Sent: Thursday, October 30, 2008 3:42 AM To: [email protected] Subject: Re: WLM Service Class "Ted MacNEIL" <[EMAIL PROTECTED]> wrote in message news:<310343960-1225328835-cardhu_decombobulator_blackberry.rim.net-1542 [EMAIL PROTECTED]>... > >At the request of management, we created a "special" job class that runs a service class which has an importance of 1, has CPU critical turned on and also has a velocity of 90%. This is running higher than CICS. We have started meeting our goal again, but they don't want to see any jobs running in this special class having any CPU delays. > > I'm glad I don't work at your shop: > 1. Management should be telling you to solve the problem. Not how to solve it! > 2. Putting batch higher than online can screw online response. > 3. 90% is almost impossible to meet. The WLM will give up on an S/C if it can't 'help' it. > 4. CPU delay is not necessarily a problem. Think of I/O-bound jobs. > 5. Yes, you can do more, but you've already over-killed. > - > Too busy driving to stop for gas! > > 6. Tell management to buy more capacity for this 1 hour per day they > have extreme demands. Hopefully this will make them *think*. > Kees. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
