Luis, See previous discussion in thread. If you have multiple target zones, it makes no difference whether they are in a single CSI or multiple CSIs, or whether the target zones are related, or share a global zone. If both target zones have elements with the same names or potentially could have elements with the same name, AND you intend to, or might unintentionally, apply maintenance to both of these zones, you *DO NOT* want to share the SCDS, LTS, STS, and MTS datasets that IBM recommends be unique for each target zone, or you run the risk of compromising the integrity of the software in the zones! This is not a matter of what is "convenient"!
Sharing these data sets is only conditionally safe when you KNOW for certain the above conditions cannot occur. Otherwise sharing these datasets across multiple target zones is like taking a loaded gun, pointing it at your foot, and hoping you miss. JC Ewing On 03/16/2010 10:21 AM, Luis Andrade wrote: > Although IBM has some recommendations on it, the SMP/E environment is > flexible and allows each one to have it defined as you see more convenient. > > We can have at least three possible scenarios for it: > > * Single-CSI structure > * Multiple-CSI structure > * consolidate previous with a Master CSI > > Anyway, you need to have, at least, these CSI data sets: > > * SMPCSI (CSI) > SMP/E uses CSIs to keep records of the system. > > * SMPPTS (PTS) > The PTS data set is used as temporary storage for SYSMODs. > It contains one member for each SYSMOD received. > Each member is called a modification control statement (MCS) entry and is an > exact copy of the SYSMOD as it was received from the SMPPTFIN data set. > The name of an MCS entry matches the ID of the SYSMOD it contains. > Generally, the MCS entries are kept on the PTS until the SYSMOD is accepted; > then, under normal processing, they are deleted. > > * SMPSCDS (SCDS) > The SCDS data set contains backup copies of target-zone entries that are > modified during APPLY processing. > These backup copies are made before the entries are > (1) changed by inline JCLIN, a ++MOVE MCS, or a ++RENAME MCS or > (2) deleted by an element MCS with the DELETE operand. > The backup copies are used during RESTORE processing to return the entries > to the way they were before APPLY processing. > > All the others are required under certain circunstances: > > – The SMPLTS data set is required when processing load modules with CALLLIBS. > – An SMPMTS DD statement is required for changes to macros that do not > reside in a target library. > – An SMPSTS DD statement is required for changes to source code that does > not reside in a target library. > > I see no problem having only one set of these SMP data sets, as long as you > know when and where you must use them. > Keep in mind that is better to have those defined in DDDEFs for each zone > with enough space for all needs, even if they are the same for all zones. > > In "SMP/E Commands" you can check for considerations when using Zone or Data > sets sharing for each command. > > If you really want to consolidate your SMP/E enviromnent consider using > ZONErelated functions, like ZONEEXPORT, ZONEMERGE and ZONEIMPORT. That way > you'll don't need to re-RECEIVE anything. > ... -- 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
