HI, The new code make big improvement in running jobs which use large block size or many CCWs in one IO. For example, I see big improvement running tape BACKUP and RESTORE using DFDSS. This improvement related to my value set in TCP MF parameters, so if your parameter set OK, your improvement will be less than my. MFNetDisk sets the right option for its TCP socket. MFNetDisk set the size of TCP buffers to what MFNetDisk uses for stacked CCW. No, user action required, because the set value is done from the MFNetDisk code and only for MFNetDisk sockets. I think that this is important fix. I expect users feedbacks about their new performance using this fix.
Thanks, Shai ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
