The IBM mainframe is by far the defacto standard as the enterprise computing platform for systems of record. These are the systems that enterprises literally run their businesses on. While the exact percentage varies by the source consulted, consensus is that 70-80% of the world’s business data runs on the mainframe. In other words, the global economy runs on z Systems. The latest IBM mainframe is the IBM z Systems z13.
Enterprises running on IBM z13 demand the highest levels of performance, scalability, and availability, particularly for their data storage subsystems. Today, longtime Brocade OEM partner EMC announced several major innovations with the EMC VMAX and EMC Disk Library for mainframe (DLm) that revolutionize mainframe environments. These innovations from EMC, combined with EMC Connectrix B-series Gen 5 FICON SAN platforms can provide IBM z13 end users with extremely high performance, low latency, and “6 9’s” levels of reliability for their mission critical workloads.
It’s All About Performance
A computing resource that is waiting to be used, especially a processor, is a wasted resource. Recall Computer Measurement Group (CMG) legend Dr. Thomas Bell’s famous quote “All CPUs wait at the same speed.
The VMAX3 and the new VMAX All Flash DASD arrays now offer twice the processing power of previous EMC DASD arrays for mainframe customers. The VMAX3 uses what EMC calls its Dynamic Virtual Matrix Architecture with all new processing engines. The VMAX3 engines are third generation Intel multi-core processors based on the Ivy Bridge architecture, and they include up to 2 TBs of cache memory (a maximum of 16 TB per DASD array) along with front end connectivity and back end SAS connectivity through two fully redundant director boards. VMAX3’s Dynamic Virtual Matrix is highly scalable as well: it can scale up to eight VMAX3 engines as a single system with fully shared connectivity, processing and capacity resources. Each engine supports up to 48 CPU cores. Other significant performance enhancing technologies in each VMAX3 include Infiniband 56Gb/s interconnect technology, PCI Gen 3 I/O technology, and native 6Gb/s SAS drive infrastructure.
With the FICON Express16S generation of features, IBM added Forward Error Correction (FEC) capabilities to the Fibre Channel link protocol on z Systems.VMAX3 and VMAX All Flash offer 16Gb/s FICON connectivity as well as support for FEC. FEC allows FICON Express16S channels to operate at higher speeds, over longer distances, with reduced power and higher throughput, while retaining the same reliability and robustness that FICON channels have traditionally been known for. When running at 16 Gbps link speeds, with FEC end users should see fewer I/O errors, thus easing the transition to the new, high-speed link technologies and reducing the potential impact to production workloads by I/O errors. Up to 11 bit errors out of every 2112 bits per block are corrected and this significantly improves reliability (it is the equivalent result to adding 2.5 db of signal strength, effectively doubling the signal’s link loss budget). FEC is a significant improvement for both performance and availability/reliability in FICON environments. Ideally, implementations will be end to end FEC capable.
The VMAX All Flash array takes this performance to the next level, delivering millions of IOPS, sustaining under 500 microseconds of latency for read/write workloads while supporting up to 150 GB/s of bandwidth. The VMAX All Flash is designed to support the most mission-critical systems and applications with six-nines availability and unprecedented performance to enable consolidation of diverse workloads onto a single VMAX.
The FICON SAN Sports Car-Tire Extrapolation
I write the Enterprise Networks columns for the Enterprise Systems Media publications. In the July 2015 issue of Enterprise Tech Journal, yours truly wrote a column entitled “The z13 Sports Car-Tire Extrapolation.” The principles I discussed there clearly also apply when upgrading your DASD arrays. In a high performance z13 environment, the FICON SAN is the equivalent to the tires of a high performance sports car such as the Chevrolet Corvette. When you buy a Corvette, you don’t re-use the tires you bought with the car you are replacing. You would not get the performance out of that Corvette, (and in all likelihood, the first time you corner fast you’ll roll the car). Likewise, when upgrading to z13 and the new EMC VMAX mainframe platforms, if you don’t upgrade the FICON SAN (your tires), you will not get the performance you are seeking.
To optimize the environment from end-to-end, the all new VMAX DASD arrays, should be deployed with the EMC Connectrix B-series Gen 5 Fibre Channel platform, which are purpose built to meet the demands of a highly virtualized environments optimized with flash. These virtualized environments need a network that can handle nonstop relentless demand to meet performance requirements. The EMC Connectrix B-series Gen 5 Fibre Channel fabrics maximize system performance, scalability, manageability and availability of all-flash storage and together with the EMC VMAX arrays, enable applications to perform tasks significantly faster. The EMC Connectrix B-series platform not only delivers the performance needed to meet SLAs but also provides deep visibility into the physical and virtual traffic behaviors, automates monitoring, and simplifies diagnostic and management capabilities with Fabric Vision technology. Together, the VMAX management tools and Fabric Vision technology allow FICON SAN administrators to realize considerable time savings, along with a comprehensive and holistic view of their virtual and physical storage environment.
The trio of IBM z13, EMC VMAX, and Brocade Gen 5 FICON is a highly synergistic combination of high performance enterprise platforms. Combining these high performance platforms together at the core of your mission critical architecture is a winning trio along the lines of The Three Tenors, The Three Musketeers, or my personal all time favorite trio: Kirk, Spock and McCoy from Star Trek.