MVS Limits Increased for Job Classes A, B, C, E, and F
Recently several users have requested changes in the limits on some of the services provided at the NIH Computer Center. As a result, the Computing Facilities Branch of DCRT has reviewed the various limits and has made the following changes, effective immediately.
For batch jobs run on the MVS system, changes have been made to the maximum and the default CPU time limits for some job classes:
Current Limits New Limits
Job Max/Default Max/Default
Class (CPU Seconds) (CPU Seconds)
A 10/10 unchanged
B 75/75 100/100
C 200000/100 20000/150*
E 150/150 20000/150*
F 9999/200 unchanged
* Note: Job class C and E users should be careful when specifying
large CPU limits, since a job that goes into a loop could be very expensive.
(See the NIH Computer Center User's Guide for the policy on refunds.)
Also, the limit for punched output for all batch jobs has been increased from 300,000 cards to 4,000,000 cards. This limit applies only to remote sites since jobs may not generate (nor read) punched cards at the central facility. Punched card output at remote sites is generally used as a method of transferring files rather than actually producing punched output.
Time-Out Interval Increased
In addition, the time-out interval for interactive WYLBUR and TSO (including TSO/DB2) sessions has been increased from 2 hours to 4 hours. Warnings will be issued after 3 hours 50 minutes of inactivity and again at 3 hours 55 minutes of inactivity. If there is still no activity after 4 hours, the session will be logged off.
Jobs that have been awaiting print or punch (or on SPOOL for any other reason) more than two weeks will be purged. Some users have directed output to remote sites that are rarely used, resulting in a number of jobs remaining in the system for considerable periods of time.
Interface 196 (June 15, 1996)