For your VB data set, is the LRECL less than 256 (e.g.: 255 or less?). From "IP User's Guide and Commands v1.13 (SC31-8780-11)" it states on page 98:
Note: ISPFStats is ignored for sequential data sets. Also, the record format must be either variable or fixed, and the record length must be less than 256. ----------------------------------------------------------------------------------- Transferring PDS member to PDS member in block mode or in compress mode differs in behavior from transferring in stream mode. If the user wants to preserve the statistics of a PDS member that already has the statistics and have the same statistics copied over to the target PDS member, transferring in block mode or in compress mode is required. Al Nims Systems Admin/Programmer 3 Information Technology University of Florida (352) 273-1298 -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Paul Gilmartin Sent: Tuesday, September 30, 2014 12:29 PM To: [email protected] Subject: Re: Any standard IBM tool to set ISPF statistics in a batch step? On Tue, 30 Sep 2014 11:54:54 -0400, Farley, Peter x23353 wrote: > >The motivation for this is the sad fact that the IBM FTP server does not set >statistics for RECFM=VB PDS/E members even when the SITE ISPFSTATS toggle is >set. It only does it for RECFM=FB PDS/E members, even at the z/OS V2.1 level. > WTF!? And why? I hadn't noticed that. Is it because it's easier to count lines if RECFM=FB (count bytes and divide by LRECL)? Or because revision level is/isn't available as the last two characters of the sequence numbers (shouldn't matter). Is this documented? Does NFS have the same behavior? I'll need to test. I know that "touch" via NFS fails to update the time stamp. Can the programmer enable/disable stats with NFS? If the data originate from a z/OS (or CMS) system, it might be better to use TRANSMIT (SENDFILE)/RECEIVE. If you want to copy to/from a UNIX (etc.) system and replicate timestamps, you're probably SOL. (TRANSMIT/RECEIVE may have a 1-hour skew if a Daylight Saving boundary (or the Equator) is between. They could have done better.) -- gil ---------------------------------------------------------------------- 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
