The CPU-cost associated with STOPX37 shows up as INIT CPU-TIME, for each initialization/active-step allocation - with significant user/admin flexibility, of course at a cost of CPU-consumption.
Along similar cause/effect to some degree/level, SMS DATACLAS functionality (less flexible than STOPX37) using SPACE CONSTRAINT RELIEF (SCR) along with DVC (dynamic ADDVOL requests, if/where possible). Scott Barry SBBTech LLC On Tue, 23 Jun 2026 21:58:17 -0500, Paul Feller <[email protected]> wrote: >Joe, the data in the dataset is not affected. The executing program does not >know that the X37 software stepped in to do the recovery process. If I recall >correctly with both of the different X37 software I worked with over the years >there are rules you can setup to try to handle things. > >Now these types of software do have limits. As an example they can't help if >there is not any DASD space to extend a dataset or you have reached the limit >for how many volumes a dataset can extend across. > > >Paul > >-----Original Message----- >From: IBM Mainframe Discussion List <[email protected]> On Behalf Of >Joseph Reichman >Sent: Tuesday, June 23, 2026 09:01 PM >To: [email protected] >Subject: Re: Automating fix for B37 from a recovery program > >So when get a x 37 type error what happens to the data that’s already written >to the file is it intact > > >Can you close the dataset reallocate and continue > >Doesn’t sound like it but I guess I can ask > > > > >Joe Reichman > >On Tue, Jun 23, 2026 at 9:57 PM Paul Feller < >[email protected]> wrote: > >> As Steve said that is what X37 type software does. The X37 software >> steps in to determine if it should try to help recover from a X37 >> error. Now if that type of software is not in the cards, then maybe >> looking at how datasets are allocated might be a way to help reduce X37 >> errors. >> >> If SMS is in the picture, you could look at some type of general rule >> around volume count. The down side to that is it takes up room in the >> TIOT below the line. If you open/allocate lots of datasets you might >> run out of room in the TIOT. If you are pushing all TIOT stuff above >> the line then space may not be an issue. >> >> Keeping with SMS maybe look at dataset allocation stuff. It comes >> down to the size of the primary and secondary allocation. Maybe some >> datasets are better off with a large primary and small secondary. >> While other datasets might work better with a small primary and large >> secondary. >> >> As far as I know SMS changes really only come into play during dataset >> allocation. Will not help with datasets already allocated. >> >> Paul >> >> -----Original Message----- >> From: IBM Mainframe Discussion List <[email protected]> On >> Behalf Of Steve Beaver >> Sent: Tuesday, June 23, 2026 07:45 PM >> To: [email protected] >> Subject: Re: Automating fix for B37 from a recovery program >> >> What you’re describing and wanting to do is exactly how the product >> stop X >> 37 works >> >> >> >> Sent from my iPhone >> >> No one said I could type with one thumb >> >> > On Jun 23, 2026, at 19:39, Joseph Reichman < >> [email protected]> wrote: >> > >> > I was able to verify what Kurtz Q told me earlier that SDWAGR02 >> > Points to the dcb in a recovery program >> > >> > That being said is it possible I guess via SVC 99 to fix I.e >> > reallocate a dataset from a recovery program >> > >> > Thanks >> > -------------------------------------------------------------------- >> > -- For IBM-MAIN subscribe / signoff / archive access instructions, >> > send email to [email protected] with the message: INFO >> > IBM-MAIN >> >> ---------------------------------------------------------------------- >> For IBM-MAIN subscribe / signoff / archive access instructions, send >> email to [email protected] with the message: INFO IBM-MAIN >> >> ---------------------------------------------------------------------- >> For IBM-MAIN subscribe / signoff / archive access instructions, send >> email to [email protected] with the message: INFO IBM-MAIN >> > >---------------------------------------------------------------------- >For IBM-MAIN subscribe / signoff / archive access instructions, send email to >[email protected] with the message: INFO IBM-MAIN > >---------------------------------------------------------------------- >For IBM-MAIN subscribe / signoff / archive access instructions, >send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
