This message group displays tuning statistics for a multipath channel-to-channel adapter. Messages IST924I - IST1570I are repeated for each subchannel address.

IST1230I TIME     = time     DATE     = date     ID  = id
IST1231I IPDU     = ipdu     OPDU     = opdu
IST1569I INLP     = inlp     ONLP     = onlp
IST1232I TSWEEP   = tsweep   QSWEEP   = qsweep
IST924I  ----------------------------------------------------------
IST1233I DEV      = dev      DIR      = dir
IST1234I BSIZE    = bsize    MAXBYTES = maxbytes
IST1235I SIO      = sio      SLOWDOWN = slowdown
IST1236I BYTECNTO = bytecnto BYTECNT  = bytecnt
IST1570I NBYTECTO = nbytecto NBYTECT  = nbytect
IST314I END

IST1230I TIME     = time     DATE     = date     ID  = id
IST1231I IPDU     = ipdu     OPDU     = opdu
IST1569I INLP     = inlp     ONLP     = onlp
IST1232I TSWEEP   = tsweep   QSWEEP   = qsweep
IST924I  ----------------------------------------------------------
IST1233I DEV      = dev      DIR      = dir
IST1234I BSIZE    = bsize    MAXBYTES = maxbytes
IST1235I SIO      = sio      SLOWDOWN = slowdown
IST1236I BYTECNTO = bytecnto BYTECNT  = bytecnt
IST1570I NBYTECTO = nbytecto NBYTECT  = nbytect
IST924I  ----------------------------------------------------------
IST1233I DEV      = dev      DIR      = RD/x (queue_type)
[IST1719I PCIREALO = pcirealo PCIREAL  = pcireal]
[IST1720I PCIVIRTO = pcivirto PCIVIRT  = pcivirt]
[IST1750I PCITHRSO = pcithrso PCITHRSH = pcithrsh]
[IST1751I PCIUNPRO = pciunpro PCIUNPRD = pciunprd]
[IST2316I EARLYINO = earlyino EARLYINT = earlyint]
[IST2317I ULPRETUO = ulpretuo ULPRETU  = ulpretu]
[IST1752I RPROCDEO = rprocdeo RPROCDEF = rprocdef]
[IST1753I RREPLDEO = rrepldeo RREPLDEF = rrepldef]
IST1754I NOREADSO = noreadso NOREADS  = noreads
IST1721I SBALCNTO = sbalcnto SBALCNT  = sbalcnt
IST1722I PACKCNTO = packcnto PACKCNT  = packcnt
IST2185I FRINVCTO = frinvcto FRINVCT  = frinvct
IST1236I BYTECNTO = bytecnto BYTECNT  = bytecnt
IST1810I PKTIQDO  = pktiqdo  PKTIQD   = pktiqd
IST1811I BYTIQDO  = bytiqdo  BYTIQD   = bytiqd
IST924I  ----------------------------------------------------------
IST1233I DEV      = dev      DIR      = WR/1 
IST1755I SBALMAX  = sbalmax  SBALAVG  = sbalavg
IST1756I QDPTHMAX = qdpthmax QDPTHAVG = qdpthavg
IST1723I SIGACNTO = sigacnto SIGACNT  = sigacnt
IST1721I SBALCNTO = sbalcnto SBALCNT  = sbalcnt
IST1722I PACKCNTO = packcnto PACKCNT  = packcnt
IST2242I SIGMCNTO = sigmcnto SIGMCNT  = sigmcnt
IST1236I BYTECNTO = bytecnto BYTECNT  = bytecnt
IST1810I PKTIQDO  = pktiqdo  PKTIQD   = pktiqd
IST1811I BYTIQDO  = bytiqdo  BYTIQD   = bytiqd
IST314I END

Many of the messages in this group contain both a count and an overflow. Both the count and overflow are maintained in unsigned 32-bit variables (unless otherwise indicated). Since an unsigned 32-bit variable can only contain a value up to and including 4294967295 ('FFFFFFFF'X), the variable will wrap through 0 if an increment results in exceeding this value. When such a wrap occurs, the overflow is incremented by 1. Therefore, the total count is determined as follows:

            Total = (overflow * 4294967296) + count

In the message text:

date and time

The date and time values specify when the record was reported. See DATE and TIME formats for information about the date and time values.

id

The name of the link for which tuning statistics are being recorded. The link name is from one of the following sources:

• The name specified on the LINE definition statement in the associated channel-attached major node.
• The name specified on the TRLE definition statement in the TRL definition deck.
• An internally generated TRLE name.

In the message text:

ipdu

The total number of inbound PDUs received.

opdu

The total number of outbound PDUs transmitted.

In the message text:

tsweep

The number of sweeps initiated as a result of a timeout. A sweep is a special data block that is exchanged with the adjacent host to verify that data has not been lost. Only the host with the highest subarea number will initiate timer sweeps.

qsweep

The number of sweeps initiated as a result of excessive receive queue depth.

• Receive queue depth represents the number of bytes of data waiting to be passed to the next layer.
• Excessive receive queue depth usually indicates a problem with reserialization of the data segment. Possible causes are a mismatch in the speed of the read subchannels or a lack of available I/O buffers.

In the message text:

dev

The hexadecimal subchannel address of the device for which tuning statistics are being recorded. It corresponds to one of the addresses coded on the READ, WRITE, or DATAPATH statement on the LINE or TRLE definition statement.

dir

The direction of this resource. Possible values are:

• READ
• WRITE
• RD/x (queue_type)
  • The x value is the QDIO read input queue number for this QDIO data device.
  • The queue_type value is the queue type. The queue_type value can be a maximum of 8 characters. Possible values include PRIMARY, BULKDATA, SYSDIST or N/A. The N/A value indicates that the queue is initialized but is not currently in use by the TCP/IP stack.
• WR/x
  • The x value is the QDIO write priority level.

In the message text:

bsize

The maximum buffer size supported by this device.

maxbytes

The number of bytes used in the largest channel program. This field provides information about the utilization or packing of data into the transmit or receive buffer. As this number approaches bsize, this indicates that maximum instantaneous utilization of the device's buffer has occurred.

In the message text:

sio

The number of start I/O operations counted for the subchannel. This number is reset each time VTAM® reports tuning statistics and is expressed in decimal. The value of sio is never larger than 65535. If sio is 65535, its value is reset to 0 when the next start I/O operation takes place.

slowdown

The number of times slowdown mode has been entered. If slowdown is incrementing, this indicates a lack of available I/O buffers.

• If DIR = READ in message IST1233I, slowdown is incremented every time the channel program cannot be reinitiated immediately because of a lack of I/O buffers to unpack the inbound data.
• If DIR = WRITE in message IST1233I, slowdown is **N/A**.

In the message text:

bytecnto

The PDU byte count overflow.

bytecnt

The byte count. This value represents the accumulated number of bytes of PDUs transmitted on the WRITE subchannel or received on the READ subchannel.

In the message text:

inlp

The total number of inbound NLPs received.

onlp

The total number of outbound NLPs transmitted.

In the message text:

nbytecto

The NLP byte count overflow.

nbytect

The NLP byte count. This value represents the accumulated number of bytes of NLPs transmitted on the WRITE subchannel or received on the READ subchannel.

In the message text:

pcirealo

The real PCI (Program Controlled Interrupt) overflow.

pcireal

The real PCI count. This value represents the accumulated number of real PCI interrupts fielded by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

A real PCI is an execution of the QDIO Program-Controlled Interrupt Exit as a result of a call from the system interrupt handler. The higher the ratio of real PCI to virtual PCI, the less successful QDIO is at avoiding the overhead of the system interrupt handler.

In the message text:

pcivirto

The virtual PCI overflow.

pcivirt

The virtual PCI count. This value represents the accumulated number of virtual PCI interrupts fielded by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

A virtual PCI is an execution of the QDIO Program-Controlled Interrupt Exit as a result of a call from the QDIO device driver. The higher the ratio of virtual PCI to real PCI, the more successful QDIO is at avoiding the overhead of the system interrupt handler.

IST1721I

• sbalcnto is the storage block address list (SBAL) count overflow.
• sbalcnt is the SBAL count. This value represents the accumulated number of SBALs used for I/O on the QDIO WRITE priority level or the QDIO READ data transfer point.

  An SBAL serves a similar function in QDIO that an indirect-data-address word (IDAW) list serves in traditional I/O.

IST1722I

• packcnto is packet count overflow.
• packcnt is packet count. This value represents the accumulated number of packets transmitted on the QDIO WRITE priority level or received on the QDIO READ data transfer point.

  A packet is a single unit of data as presented to the QDIO device driver (for example, a datagram or TCP packet).

IST1723I

• sigacnto is SIGA (Signal Adapter) count overflow.
• sigacnt is SIGA count. This value represents the accumulated number of SIGA instructions issued for the QDIO WRITE priority level.

  SIGA tells the QDIO adapter that data is ready to be written. This count will not directly correlate with the data rate as QDIO employs a tack-in mechanism similar to Seldom Ending Channel Program.

IST1750I

• pcithrso is threshold PCI overflow.
• pcithrsh is threshold PCI count. This value represents the accumulated number of threshold PCI interrupts fielded by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

  A threshold PCI is a real PCI generated by the adapter because one of the threshold conditions which controls the PCI processing was met.

  A count of zero indicates that the QDIO device driver is providing sufficient resources to keep pace with the inbound data stream from the adapter.

IST1751I

• pciunpro is unproductive PCI overflow.
• pciunprd is unproductive PCI count. This value represents the accumulated number of unproductive PCI interrupts fielded by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

  An unproductive PCI is a real PCI where the QDIO Program-Controlled Interrupt Exit failed to find any completed reads. An unproductive PCI will occur when a virtual PCI causes the processing of read completions for which a real PCI is pending. An unproductive PCI indicates that the system interrupt handler overhead was incurred unnecessarily.

IST1752I

• rprocdeo is read processing deferral overflow.
• rprocdef is read processing deferral count. This value represents the accumulated number of read processing deferrals by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

  A read processing deferral occurs when the QDIO Program-Controlled Interrupt Exit must defer a read completion because a control block cannot be obtained to represent the inbound data. Read processing deferrals (along with read replenishment deferrals) might cause the NOREADS count to be nonzero.

IST1753I

• rrepldeo is read replenishment deferral overflow.
• rrepldef is read replenishment deferral count. This value represents the accumulated number of read replenishment deferrals by the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

  A read replenishment deferral occurs when the QDIO Program-Controlled Interrupt Exit does not have enough available read buffers to tack-in a new read. Read replenishment deferrals (along with read processing deferrals) might cause the NOREADS count to be nonzero.

IST1754I

• noreadso is reads exhausted overflow.
• noreads is reads exhausted count. This value represents the accumulated number of times reads were exhausted on entry to the QDIO Program-Controlled Interrupt Exit for the QDIO READ data transfer point.

  This value is incremented by 1 each time the QDIO Program-Controlled Interrupt Exit is invoked and it detects that all the read buffers are full (the adapter has no place to move additional inbound data). A zero total is preferred because lack of read buffers might result in the adapter discarding inbound data. Examination of the read processing and read replenishment deferral counts might indicate the reason the QDIO device driver is not providing sufficient resources to accept the inbound data from the adapter.

IST1755I

• sbalmax is the maximum number of active SBALs at the completion of the write initiation process for the QDIO WRITE priority level. This value will be in the range 0-128 (0 meaning the priority level had no outbound activity in the interval and 128 meaning that at one point in the interval ALL the SBALs for the priority level were active).
• sbalavg is the average number of active SBALs at the completion of the write initiation process for the QDIO WRITE priority level. This value will be in the range 0-128 (0 meaning the priority level had no outbound activity in the interval and 128 meaning that every time the QDIO write initiator completed, all 128 SBALs were active).

IST1756I

• qdpthmax is the maximum number of work elements left on the outbound work queue at the completion of the write initiation process for the QDIO WRITE priority level. This value will be a number in the range 0–54 or a character constant >254.
• qdpthavg is the average number of work elements left on the outbound work queue at the completion of the write initiation process of the QDIO WRITE priority level. This value will be a number in the range 0–254 or a character constant >254.

  A nonzero value for either of these fields suggests the adapter is not accepting outbound data as fast as the device driver is presenting it.

IST1810I

• pktiqdo is the IQDIO packet count overflow.
• pktiqd is the IQDIO packet count. If this message is for a QDIO or IQDIO DATAPATH device with a direction of read, the value represents the accumulated number of packets received on this particular QDIO or IQDIO READ data transfer point and subsequently routed outbound using IQDIO routing or QDIO Accelerator. If this message is for a QDIO or IQDIO DATAPATH device with a direction of WR, the value represents the accumulated number of packets transmitted on the write priority level for this device after being received inbound and subsequently routed outbound using IQDIO routing or QDIO Accelerator.
• A packet is a single unit of data as presented to the QDIO device driver (for example, a datagram or TCP packet).

IST1811I

• bytiqdo is the IQDIO byte count overflow.
• bytiqd is the IQDIO byte count. If this message is for a QDIO or IQDIO DATAPATH device with a direction of read, the value represents the accumulated number of bytes received on this particular QDIO or IQDIO READ data transfer point and subsequently routed outbound using IQDIO routing or QDIO Accelerator. If this message is for a QDIO or IQDIO DATAPATH device with a direction of WR, the value represents the accumulated number of bytes transmitted on the write priority level for this device after being received inbound and subsequently routed outbound using IQDIO routing or QDIO Accelerator.

IST2185I

• frinvcto is the frame invalidation error count overflow.
• frinvct is the frame invalidation error count. This message applies to inbound data received on a QDIO DATAPATH device. The value represents the accumulated number of frame invalidation packets that were received and discarded. A frame invalidation packet is an inbound packet marked as invalid by OSA-Express3 or later.
• Result: These fields are 0 for OSA-Express, OSA-Express2 or later, and iQDIO ports.
• For detailed instructions about setting up an OSA-Express feature, see the zEnterprise System and System z10 OSA-Express Customer's Guide and Reference.

IST2242I

• sigmcnto is the iQDIO (HiperSockets) multiple write SIGA count overflow.
• sigmcnt is the iQDIO (HiperSockets) multiple write SIGA count. This value is the accumulated number of signal adapter (SIGA) multiple write instructions issued for the IQDIO WRITE priority level.
• SIGA multiple write tells the IQDIO device that multiple storage block address lists (SBALs) with data are ready to be written. This count does not directly correlate with the data rate because the number of SBALs represented by each SIGA multiple write might vary.

IST2316I

• earlyino is the early interrupt count overflow.
• earlyint is the early interrupt count. This message applies to inbound data that is received on a QDIO DATAPATH device. The value represents the number of times that an OSA-Express feature that is operating in optimized latency mode signaled the host with an interrupt that indicates that no data was currently available to process, but that data that is destined to this host was received.
• Result: These fields are both 0 for any QDIO DATAPATH device that is not operating in optimized latency mode.

IST2317I

• ulpretuo is the Upper-Layer Protocol (ULP) return with no data available count overflow.
• ulpretu is the ULP return with no data available count. This message applies to inbound data received on a QDIO DATAPATH device. The value represents the number of times that VTAM presented data to the ULP from an OSA-Express feature that is operating in optimized latency mode, and when VTAM received the ULP return, it found that there was no more data to process.
• Result: These fields are both 0 for any QDIO DATAPATH device that is not operating in optimized latency mode.