FICON Express2 Channel Performance Version 1.0 - IBM

More documents

Recommendations

Info

FICON Express2 Channel Performance Version 1.0 Page 37 paths 01, 02, 04 and 05 are responding on average in about 0.5ms. In this case, if no errors were made in the IOCDS, then some “tuning” of the configuration is necessary to reduce these CU busies and achieve better response time results. Contention due to CU busies results in higher than normal PEND times and contributes to a higher than normal average number of open exchanges for this LCU. The solution to this is to identify the source of contention at the CU ports connected to channel paths 03 and 06 in this example and fix it. I/O Q U E U I N G A C T I V I T Y from a configuration with FICON CU port contention causing high pend times LCU CONTROL UNITS 0222 1000 AVG AVG CHAN CHPID % DP % CU CUB CMR PATHS TAKEN BUSY BUSY DLY DLY 01 89.256 0.00 0.00 0.0 0.5 02 86.348 0.00 0.00 0.0 0.5 03 1.908 0.00 15.08 0.0 203 04 89.644 0.00 0.00 0.0 0.5 05 86.055 0.00 0.00 0.0 0.5 06 2.132 0.00 13.95 0.0 207 * 355.34 0.00 0.19 0.0 2.9 note: CU busies and high CMR delay times is NOT normal for native FICON...indicates CU port contention Figure 29 For FICON channels it is also possible to estimate the average number of bytes transferred per SSCH by dividing the MB/sec of a FICON channel from the Channel Path Activity report by the total SSCH/sec processed by a FICON channel from the I/O Queuing Activity report. The total SSCH/sec processed by a FICON channel can be determined by adding up all of the “CHPID taken” fields on the I/O Queuing Activity report for each LCU that a single FICON channel is connected to. If the average data transfer sizes of your channel programs are greater than 27K bytes, then your workload is most likely pushing the channel and CU port buses and links to higher levels of utilization than other resources and you should focus on the MB/sec fields on your RMF Channel Activity and FICON Director Activity Reports and compare these to the maximum capability of the FICON channels, CU ports and links used in your configuration. In summary, the basics of performance analysis do not change with a FICON configuration versus an ESCON configuration. In both environments, an appropriate technique to use is to first calculate I/O intensities, which equals I/O rate multiplied by response times. This analysis can be done at a device volume level, an LCU level, a physical CU box level or for a group of channels. The parts of the total I/O configuration that have the highest I/O
FICON Express2 Channel Performance Version 1.0 Page 38 intensities are the “hot spots” of the configuration. These are the areas where configuration tuning has the potential for yielding the highest benefit. As explained above, the individual components of response time (IOSQ, DISC, PEND and CONN) will tell you where you should focus your efforts. The average open exchange calculation is a subset of the I/O intensity calculation that uses the DISC + CONN + CMR components of response time. Except in cases of extremely low control unit cache hit ratios, the open exchange limit is not the cause of high values of average open exchanges. Instead high values for average open exchanges are most likely the result of driving either the channels or the control unit to high levels of utilization. Tuning efforts need to be focused on the appropriate areas based on the DISC, CONN and CMR components of workload response times. If the FICON channel processor and bus utilizations as reported on the RMF Channel Activity report and link utilizations calculated from the MB/sec info are less than 50%, then the tuning efforts need to focus on the control units in the configuration. The basic architecture and design differences between FICON and ESCON resulted in many changes to the performance data that appear on RMF reports. Additional information in the form of FICON processor and bus utilizations, READ and WRITE MB/sec, AVG FRAME SIZE and AVG CMR DLY is provided to help analyze the multiplexing capability of FICON. Since ESCON is only capable of executing one I/O operation at a time, RMF reports the time that the entire CHPID path is busy for ESCON channel utilization. With FICON, we must consider the individual components of the total CHPID path such as the FICON channel processor and bus, the fibre link, the director destination port and the control unit port adapter microprocessor, bus and link. The charts and examples provided in this paper should help guide you in assessing the maximum capability of FICON Express2 channels for your workload.
Page 1 and 2: April 2005 FICON Express2 Channel P
Page 3 and 4: FICON Express2 Channel Performance
Page 37: FICON Express2 Channel Performance
Page 41: FICON Express2 Channel Performance

FICON Express2 Channel Performance Version 1.0 - IBM

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?