In a message dated 12/5/2007 3:16:56 P.M. SA Western Standard Time, [EMAIL PROTECTED] writes:
RS, a) A CA split is still costly, but having extra data buffers normally helps though (BUFND=); a CI split is almost nothing now. b) Empty CIs are not read by VSAM. Almost empty CIs cost the same for random read as for full CIs. For sequential read almost empty CIs does cost extra. You can always reorg to pack the data better. VSAM can be made much more efficient by appropriate buffering, e.g. System Managed Buffering is pretty good. VSAM splits are a fact of life, and it is an Access Method that is 30+ years old. It could be worse, we could still be stuck with ISAM!! (yes, I did work with it). Regards. Mike ------------------------------------------------------------------------------ ------------------------------------ Mike Poil Java z/OS Level 3 Service IBM United Kingdom Limited, Hursley Park, Winchester SO21 2JN Internal: 246824 External: +44 (0)1962 816824 Java debugging: http://www.ibm.com/developerworks/java/jdk/diagnosis/ ------------------------------------------------------------------------------ ------------------------------------ "R.S." <[EMAIL PROTECTED]> Sent by: IBM Mainframe Discussion List <IBM-MAIN@BAMA.UA.EDU> 05/12/2007 18:00 Please respond to IBM Mainframe Discussion List <IBM-MAIN@BAMA.UA.EDU> To IBM-MAIN@BAMA.UA.EDU cc Subject Re: KSDS CA split, no CI split Gene Hudders wrote: [...] > I sure would hate to take too many CA splits in an on-line environment > especially if the CA split requires another extent for either the data > and/or index component. That's why I wrote "not so bad". <g> Indeed CA split is painful (CI also, but the pain is CI/CA times less). However the pain is no longer because after the split data is no longer in physical sequence. That (maybe) was painful in times of SLEDs with long seek time. What is painful, it's: a) poor performance *at the split time* - you insert single LR (logical record) and VSAM moves cylinder of data. b) free space distribution. After the splits you could end up with 50% free space. Overhead because of reading almost empty CIs. Did I miss something ? Regards -- Radoslaw Skorupka Lodz, Poland -- BRE Bank SA ul. Senatorska 18 00-950 Warszawa www.brebank.pl Sd Rejonowy dla m. st. Warszawy XII Wydzia Gospodarczy Krajowego Rejestru Sdowego, nr rejestru przedsibiorców KRS 0000025237 NIP: 526-021-50-88 Wedug stanu na dzie 01.01.2007 r. kapita zakadowy BRE Banku SA (w caoci opacony) wynosi 118.064.140 z. W zwizku z realizacj warunkowego podwyszenia kapitau zakadowego, na podstawie uchwa XVI WZ z dnia 21.05.2003 r., kapita zakadowy BRE Banku SA moe ulec podwyszeniu do kwoty 118.760.528 z. Akcje w podwyszonym kapitale zakadowym bd w caoci opacone. ---------------------------------------------------------------------- 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 Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU ---------------------------------------------------------------------- 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 Hi RS: (SS=sequence set and IS=index set) Extra buffers to improve a CA split requires the data set to be defined in NSR under CICS which is not necessarily the best option. Painful is longer response times for all transactions trying to access the file during the CA split, NSR is not supported under Transaction Isolation, a NSR file locks the TCB under which it runs until the split is complete (therefore, you want to make sure you define the CO TCB (on a multi-processor) in these cases where you use NSR), the formatting of the new CA, the movement of individual CIs to new CA, the clearing of the moved CIs from the original CA, the updating of the indices (SS and IS) which may result in a CI split when the new high key is added to the IS index and there is insufficient space to accommodate it, and if this is not enough of a penalty, you could wind up not having any space to allocate the new CA which results in having to access and update the catalog and VTOC (the price is similar to that of opening the file) with the exposure that there is a volume switch which may result in a no space condition exposure if insufficient space to allocate a primary or if SMS controlled a secondary and finally the same applies to the index component when adding the new index records (SS and possibly IS if split). I am sure we are not talking milliseconds even with the new disk technology. The net is you don't want CA splits on-line. That is probably not possible all the time but the idea is to minimize the occurance. Regards, Gene **************************************Check out AOL's list of 2007's hottest products. (http://money.aol.com/special/hot-products-2007?NCID=aoltop00030000000001) ---------------------------------------------------------------------- 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
