Storage problems can be related to a specific I/O device.
Because outbound data cannot be transmitted to an I/O device until
it is accepted by that device (that is, until write processing completes),
there are scenarios in which this storage associated with this data
can accumulate at the DLC (data link control) layer. For all I/O devices
that perform real I/O that are represented by a TRLE (predefined or
dynamically built), VTAM® tracks
the outbound workload (units of work) for each device. This tracking
mechanism allows console operators to isolate this type of problem
to a specific device.
The console operator can quickly isolate a storage problem
to a specific device using the DISPLAY TRL command. If a device has exceeded
internal thresholds, message IST1800I is issued with the text **
CONGESTED **. Additional details regarding the workload for
a specific device are displayed with the DISPLAY TRL,TRLE=trlename command. VTAM displays the current, average,
and the maximum workload for each device. When the current workload
is excessive, the I/O activity for this device might be associated
with system storage shortages.
If the counts for a device reveal an excessive current
workload, additional steps are required to isolate the problem.
- Steps for a console operator
When a device is marked
as congested, further action is required to determine whether the
congestion is related to a system storage problem. If the following
steps indicate that a system storage shortage is present, it might
be necessary to obtain documentation (such as a console log and a
dump) to diagnose the congestion related to this device. This condition
might be relieved by deactivating the PU (or stopping the device for
TCP/IP).
- Review the system console for any messages related to current
storage shortage conditions.
- Issue the following VTAM display
commands:
- D NET,CSM
- D NET,BFRUSE
- D NET,STORUSE,POOL=*
Note: If applicable, also issue the TCP/IP DISPLAY command
D TCPIP,,STOR.
- Issue D NET,TRL,TRLE=trlename to obtain
more details about the device congestion. Message IST1802I displays
detailed counts of units of work for the device measured at the Data
Link Control (DLC) layer.
- Activate VTAM tuning statistics
(TNSTAT), RMF™, or other monitoring
tools to monitor this specific device.
- Display the active jobs in the system to determine whether new
work was recently started.
- Steps for a system programmer
The following steps might
be required to isolate a system storage problem that is related to
an I/O device:
- Review the network configuration related to this device or any
recent configuration changes for this system.
- Review or monitor (using the output from VTAM TNSTAT or RMF)
the network traffic related to this device. Compare the actual workload
to the I/O capacity of the hardware device.
- Determine if the congestion is related to a specific time of day,
job, application, or type of workload.
- Verify that missing interrupt handler (MIH) is enabled for the
write devices.
- Review or verify that the maintenance level for the hardware device
is current.
- Consider automating the necessary storage displays to monitor
system conditions.