zHyperLink: The Holy Grail of Mainframe I/O?

Dr. Gilbert HoutekamerBy Gilbert Houtekamer, Ph.D.zHyperLink

Now that it has become harder and harder to make the processor faster, IBM is looking for other ways to make their mainframes perform better.

This has resulted in new co-processors for compression and encryption and now also, with the z14 processor, in a new technology called zHyperLink. This new I/O connectivity aims to significantly reduce the I/O response time, while at the same time not increase the processor (CP) load.

This new technology comes with a set of promises and restrictions that will cause you to rethink the design of your storage and replication infrastructure. The days of distance limitations are back, which has big implications for synchronous replication in particular.

zHyperLink I/O is synchronous I/O, which means that the processor will wait until the I/O operation is complete. This can be more efficient than a POST and SRB, but overall little change is expected. It means that your MLC invoice will not be impacted, it might even be reduced a tiny bit. It will, however, likely impact what is reported as TCB and SRB time, which in turn may have an impact on your chargeback systems.

Our latest white paper, zHyperLink: The Holy Grail of Mainframe I/O?, will discuss:

  • The instrumentation that IBM provides in RMF and SMF records
  • IBM’s positioning of the new zHyperLink technology
  • Some considerations for installations that consider deploying it

Download the white paper, “zHyperLink: The Holy Grail of Mainframe I/O?” here.

 

zHyperlink: the holy grail of mainframe IO

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.