I can't speak for other virtual tape implementations, but for CA-Vtape you can define by tape data set name which of a fairly large number of aggregation groups should be used for the virtual volume. If you can correctly categorize the virtual volumes into aggregation groups by how many days or weeks before they expire, then it is possible to have aggregation groups where no recycle is needed because all virtual volumes on a physical tape expire within a few days of each other, freeing the physical volume.
Like HSM though, you can get so much on a single cartridge that duplexing should always be used. If Vtape externalization could occur at the same time as HSM writing to ML2 or performing backup to tape, then you are immediately at a requirement for 4 drives, even without anyone needing to read HSM or virtual tape data. J C Ewing On 11/18/2009 01:16 PM, rr14 wrote: > Hello All, > > Seeking some advice/input. > > We are in the process of moving from 3490 to 3592 tapes. > For most items we have completed and are implementing a transition to > HSM volumes and then > using HSM to archive to ML2 tapes. > > For our image copies though, the number and size would require us to > do many HSM tape > recycles. Due to the length of time of the recycles, and that they > would use all of our tape > drives we would like to come up with a different scenario for the > image copies. > > Our first thought was to put image copies on DASD, but that would > require over 250 volumes. > Our second thought was to write to DASD and then archive the contents > to tape and reuse the > volumes, and reuse the tapes every 10 days, thus only requiring 25 > volumes, but our DBA says > the data sets need to remain cataloged. > Do the image copy data sets have to remain cataloged for DB2 to be > able to use them? > > We are a small shop and do not have the CPU to compress anything, and > have a small pool of 3592 tapes, thus > cannot allow jobs to write directly to the tapes. > > Virtual Tape products seem to have the same inherent issues as using > HSM L2 processing. > > Does any one have a better idea? > > Thanks...Rob -- Joel C. Ewing, Fort Smith, AR jremoveccapsew...@acm.org ---------------------------------------------------------------------- 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
