You have hit upon the correct solution, change the Management Class. I suggest the use of PRIMARY DAYS and LEVEL1 DAYS in lieu of "generations on primary".
If the "generations on primary" setting causes the GDG to be held on primary for more that Level1 days, this could cause your situation. When the dataset is finally eligible for migration due to "generations on primary" is has exceeded Level1 days, and will thus go directly to ML2, in your case, tape. If the "generations on primary" is set to a small number or blank, the dataset may be eligible for migration in less than Level 1 days, the dataset will make an interim stop at ML1 until Level 1 days has expired and then move to ML2. After you make this change, ensure you have sufficient ML1 space to hold the additional data. BTW, I do not think you can direct some ML2 to tape and some to disk. IIRC it is all or nothing. Existing ML2 data on tape will remain there, but new ML2 data will go to the tape or disk as specified in the SETSYS command. Recycle output will go to the destination specified in SETSYS RECYCLEOUTPUT. HTH, <snip> Hi Listers, We would like to reduce the tape utilisation of HSM by adding more DASD for migration. The question is whether to add more as ML1 or add some as ML2. If we just add more as ML1 then I don't see too much benefit. Currently there is not much migration occurring ML1 to ML2 on a daily basis. Approx 75% of all migration is Primary direct to ML2 tape. I believe that this is mostly GDS datasets migrating due to "generations on primary" setting in the management class. Is anybody mixing DASD and TAPE for ML2? If so how do you control the migration? We don't want to convert all our ML2 to tape. Alternatively, is there a way to force these direct Primary to ML2 migrations to be directed to ML1 instead of ML2? <snip> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
