Greg Boyd wrote: >Unfortunately, the RMF Crypto Hardware Activity report will be focused on the >utilization of the crypto cards (the Crypto Express cards), not the CPACF. >The CPACF simply provides some additional instructions that perform crypto >operations (Cipher Message, Compute Message Digest, etc.). And just as you >can't break out the CPU utilization of a single Move Character or Shift >Logical instruction, you can't break out the CPU utilization of these >instructions either.
Thanks. This is what I also discovered. I was afraid that I missed something, but ... >That said, starting with the z10 machines we did make the CPU Measurement >Facility available. This facility gives you the ability to 'count' >instructions and that includes many of the crypto instructions. John Burg has >done presentations on the use of the Measurement Facility at Share and >zExpo/zTech University. You can access at least one version of the >presentation on the IBM ATS TechDocs website >(http://www.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/TC000005). Thanks for that interesting link. I will talk with my z/OS Team about that. >I should point out that the Measurement Facility captures data at the LPAR >level. In contrast, the RMF Crypto Hardware Activity report is for the entire >CEC, and thus the entire crypto card. The point being that you have to be >careful how you interpret the reports (both of them). Thanks! >John has experimented with using the data from the counters to calculate the >CPU consumption of the various instructions. If nothing else, you should >probably monitor these counters (at least for the crypto instructions) so you >can develop trend lines and monitor the use of crypto at that level. Greg, many thanks for your kind reply. I am humbled by your great kindness! ;-) Groete / Greetings Elardus Engelbrecht ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
