Yes, exactly. There are entries in the JES2 manuals. But, and big but, everything depends on How Many JOEs - if you currently have 10,000 and want to go to 100,000 BERTs - if you have 10,000 and want to support 80,000 jobs - do you have enough? JQEs - If you currently have 10,000 and want to go to 80,000 JNUMs And other like resources.
The intent is to try anticipate what is going to be needed, not make a guess and see if JES2 allows the change. For example, Shop X has been running JES2 successfully for many decades. Now they want to increase the number of jobs that the spool can hold. Without knowing in advance if it can be supported with their current configuration, They look at the RM panel of SDSF and see what they are current using. They issue the $TOUTDEF,JOES=100000. They get the $HASP542 message saying too small, needs another 5000 4K blocks. Now this shop will need to go and update the CF Policy in JES2. 1) Find the HASP537 message at startup to show how many 4k blocks are there 2) Add that number to the 5000 K blocks 3) User CFSIZER to get new numbers. 4) Plan to update the CF Policy for the JES2 Checkpoint Structure 5) Apply to any environments as needed. For example DR system. They will need to review any physical JES2 Ckpt datasets they have to see if they are large enough If they could have done a "what if" first, made a better guess. They may have been able to update the CF Policy much earlier in preparation of the change. As well as know the relationship among BERTs, JQEs, JOEs, JNUMs. This is really not well understood anymore. A question to ponder. If you had to change a JES2 Resource, do you know if you have sufficient size to handle it? Lizette > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Carmen Vitullo > Sent: Tuesday, January 17, 2017 9:14 AM > To: [email protected] > Subject: Re: How to calculate the JES2 Checkpoint and Spool Space > > The Init and Tuning Guide has a formula for DASD sizing , reading a bit...they > want you to define the checkpoint for a secondary JES subsystem and start it > up and 'see' if its the right size. :( but there is a formula in the guide > > > > Specifying the storage of checkpoints on DASD > When you have calculated the number of 4K records for the checkpoint, use > Table 53 to determine the number of tracks or cylinders on DASD, and round up > to > the nearest whole number. > Table 53. Number of 4K records per track and cylinder for various DASD devices > DASD Device Type 3380 3390 9340 > # of 4K Records per Track 10 12 10 > # of 4K Records per Cylinder 150 180 150 > Note: > 1. Track 1 contains only the Check, Lock, and Master records, plus as many > change log records (4K each) that will fit on the remainder of the track. > Track 1 > correction is needed because track 1 is typically not filled to capacity. > 2. The change log records can overflow onto subsequent tracks. > 3. If the data set is not large enough, JES2 issues the $HASP296 message to > indicate the number of additional tracks required. (There is no penalty for > over-estimating the number.) > For example, 4K records require the following amount of space on DASD Device > Type 3380 or 3390, if the number of 4K records for type 3380 is 376 and the > number of 4K records for type 3390 is 378: > DASD Device Type 3380: > 38 tracks (376 / 10 = 37.6), or 3 cylinders (376 / 150 = 2.5) > DASD Device Type 3390: > 32 tracks (378 / 12 = 31.5), or 3 cylinders (378 / 180 = 2.1). > ----- Original Message ----- > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
