This message is the first in a group of messages that displays tuning statistics pertaining to the VTAM® operation of a channel-to-channel adapter. A complete description of the message group follows.

IST577I    TIME = time        DATE = date          ID = id
IST578I   CHNRM = chnrm      CHMAX = chmax      RDBUF = rdbuf
IST579I    ATTN = attn      TIMERS = timers     QDPTH = qdpth
IST580I  BUFCAP = bufcap       PRI = pri        SLODN = slodn
IST581I    IPIU = ipiu        OPIU = opiu      DLRMAX = dlrmax
IST1022I  WRBUF = wrbuf
IST314I  END

IST577I

• time indicates the time (in hours, minutes, seconds, and hundredths of seconds) at which the record is recorded. For example, 07431380 means that the record was recorded at the 7th hour, 43rd minute, 13th second, and 80 one-hundredths of a second of the day.
• The date value specifies when the tuning statistics report is recorded. The date is in the form yyddd, where yy is the last two digits of the numeric year and ddd is the numeric day of the year. For example, 87190 means the record is recorded on the 190th day of 1987.
• id provides the name of the link through which the tuning statistics are taken. It corresponds to the name of the LINE definition statement in the associated channel-attachment major node.

IST578I

• chnrm is the number of channel programs issued that VTAM used to send data to the node on the other side of the adapter.
  • chnrm will be greater than or equal to the number of write triggers (TIMERS + QDPTH + PRI + BUFCAP).
  • The difference between chnrm and the sum of the write triggers represents the following:
    • The number of channel programs with write data that are initiated by an attention from the other host when data was queued, but a channel program with write data could not be triggered.
  • As you increase the value of the DELAY operand for the channel-to-channel adapter, the difference between chnrm and the sum of the write triggers might be greater.
• chmax is 0 because all channel programs are the same size.
• rdbuf is the total number of input bytes transferred during the measurement period.

IST579I

• attn is the number of times a channel program is initiated because the other host has data to send. This statistic cannot be correlated with any of the other statistics that are provided; it is simply a value that indicates the number of attention interrupts.
  • When compared over an interval of time, ATTN usually does not equal the sum of TIMERS, QDPTH, BUFCAP, and PRI at the other host. VTAM counts only the first event that initiates an I/O operation, and when both hosts try to write at once, one of the hosts receives an attention that is not counted in its tuning statistics.
• timers is the number of times a channel program with write data is started because the period specified for queuing channel-to-channel PIUs has expired.
  • If session traffic is heavy, the desirable value is 0.
  • If session traffic is light, a low value rather than 0 is desirable.
  • Increasing the DELAY operand on the LINE definition statement or using transmission priority 2 might decrease the value of timers.
• qdpth is the number of times a channel program is initiated because the queue limit has been reached. This number should be higher than timers.
  Note: If DELAY=0 is specified for the channel-to-channel adapter, the TIMERS and QDPTH tuning statistics might be misleading.
  • If DELAY=0, qdpth indicates the number of channel programs that wrote data to the channel-to-channel adapter. VTAM determines the QDPTH limit based on usage except in the case of DELAY=0.
  • If DELAY=0, timers does not increment.

IST580I

• bufcap is the number of times a channel program with write data is initiated because there is enough data to fill the read buffers of the host on the other end of the channel.
  • bufcap will also be incremented if a channel program with write data is initiated due to residual PIUs left on the data queue after a channel program with write data containing a full write buffer of data has completed.
  • If bufcap is always 0, the other VTAM host has too many read buffers.
• pri is the number of times a channel program with write data is started because a high priority PIU is on the outbound channel queue; that is, the PIU is running under transmission priority 2 or is a virtual route pacing response.
  • If this number is high and there is very little transmission priority 2 traffic over this channel, the minimum virtual route window sizes are probably too small. The higher this number is in relation to the sum of TIMERS + QDPTH + BUFCAP, the less outbound coattailing occurs, and the more CPU time is used for each PIU.
• slodn indicates the number of times that this VTAM had channel programs with write data blocked by a slowdown condition in the other VTAM.

IST581I

• ipiu is the number of inbound PIUs. The average number of PIUs for each channel program can be calculated from the sending side as OPIU / (CHNRM + CHMAX).
• opiu is the number of outbound PIUs. The average number of output PIUs for each channel program with write data can be calculated as OPIU / (CHNRM + CHMAX).
• dlrmax is a decimal value that indicates the maximum number of dump-load-restart requests that were awaiting processing or were being processed at one time during the interval. This number refers to the entire domain, not to the SNA controller named in the report. The dump-load-restart subtask processes the following types of requests:
  • Dump, load, or restart of an NCP
  • Some VTAM messages to the operator that require a reply
  • Session establishment and termination processing for a local major node
  • Any I/O to a configuration restart or NODELST file.
  • This value can be used to determine the proper setting for the DLRTCB start option, which determines how many dump-load-restart requests can be processed concurrently. If DLRMAX consistently exceeds DLRTCB, it indicates that VTAM is serializing requests on the available TCBs and that performance might be affected.

IST1022I

• wrbuf is the total number of output bytes transferred during the measurement period.