W dniu 04.04.2025 o 16:20, Jousma, David pisze:
So, we are finally biting the bullet and making 1TB EAV volumes our standard
for UCB relief, etc. One “habit” my Storage team over the years had done to
mask poor dataset allocations was to make the mainly used DATACLAS
multi-volume, with a max of 59 volumes. It was before my time, but I’m 99%
sure that was done to avoid x37 abends for max extents. Now that we are
going to EAV’s, we’ll be clearing off the mod-54’s and DISNEW them. And with
a 20-1 reduction in physical volumes, there will likely be pools that had say
100 volumes, that could end up with just 5 EAV’s.
I’m not a storage guy by craft, so am wondering if there is any magic dust to
help with this other than making the owners of the poorly allocated files make
adjustments to their allocations?
That't the clou. IMHO the storage admin is "the owner" of the datasets.
Yes, he has to cooperate with DB2 guys, other administrators and
application administrators/programmers to adjust allocations.
Years ago it was storage admin duty to find suboptimal blocksizes and
change it to SDB, whenever possible, even demand changes in COBOL programs.
Yes, it is loooong process, but rush would be risky IMHO.
--
Radoslaw Skorupka
Lodz, Poland
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
