Explanation
CHECK(IBMIOS,IOS_CMRTIME_MONITOR) determined that
one or more control units had inconsistent command response (CMR)
time. Each of the control units listed in this report had at least
one channel path with an average CMR time that exceeded the THRESHOLD
value and that path was significantly higher (as defined by the RATIO
value) than the path with the lowest average CMR time for this control
unit. That is, the average CMR time for this path was at least 'x'
times the lowest average CMR time for this control unit (where 'x'
is the RATIO value).
For example, with a THRESHOLD value of
3ms and a RATIO value of 5, a control unit will be included in this
report if it had a path with an average CMR time that was the highest
among other paths to the control unit and met both the following conditions:
- The average CMR time for this path was greater than 3ms.
- The average CMR time for this path was at least 5 times greater
than the path with the lowest average CMR time for this control unit.
This may happen when there is a problem somewhere in
the fabric, such as, hardware errors, I/O mis-configurations, or congestion.
Note: If
parameters have recently been changed, this result may be based on
previously defined parameters. Newly updated parameter values will
be used the next time control unit data is gathered.
In the
message text:
- threshold
- The THRESHOLD value used by this instance of the check,
in conjunction with the RATIO value, to determine which
control units had an exception. Each of
the control units listed in this report had
at least one channel path with an average
CMR time that exceeded this THRESHOLD value
in addition to being significantly higher (as defined by
the RATIO value) than the path with the lowest
average CMR time for this control unit.
See IBM® Health Checker
for z/OS® User's
Guide for more information on the THRESHOLD
parameter.
- x
- The RATIO value used by this instance of the check,
in conjunction with the THRESHOLD value, to determine
which control units had an exception. Each control
unit listed in this report had a least one
channel path with an average CMR time that exceeded
the THRESHOLD value and was at least 'x' times
the average CMR time for the path with the lowest
average CMR time. See IBM Health
Checker for z/OS User's
Guide for more information on the RATIO parameter.
- cu
- The control unit number of a control unit that
encountered the exception. This is the control unit
number associated with the first installed path.
If '****' appears, z/OS was
unable to determine the control unit for the path.
- node
- The node descriptor of the attached subsystem in the form tttttt.mmm.nnn.pp.ssssssssssss,
where:
- tttttt
- Type number of this device.
- mmm
- Model number of the device.
- nnn
- Manufacturer identification code.
- pp
- Plant of manufacture number.
- ssssssssssss
- Sequence number.
- chpid
- A channel path (X'00' through X'FF) attached to the
control unit.
- el
- A one or two byte physical link address of the channel
path to the entry switch. If the link address defined
in the I/O configuration is one byte, it represents
the entry port address. If the link address
defined in the I/O configuration is two bytes, the
first byte is the switch address (domain) and the
second byte is the entry port address (area). If the
channel is connected directly to the control unit (point
to point configuration), a one byte link address
is shown. '..' is shown if the link address
could not be obtained for this control unit
path.
- dl
- A one or two byte address of a control unit that is used
to access the I/O device associated with the specified
channel path. If the link address defined in the
I/O configuration is one byte, it represents the
destination port address. If the link address
defined in the I/O configuration is two bytes,
the first byte is the switch address (domain) and the second
byte is the destination port address (area). If
the channel is connected directly to the control
unit (point to point configuration), a one
byte link address is shown. '..' is shown if
the link address could not be obtained for this
control unit path.
- intf
- A two byte physical identifier that represents the
control unit interface that is connected to the channel
path.
- iort
- The number of I/Os started down this channel path per
second.
- cmr
- The average command response (CMR) time for this channel
path displayed in milliseconds. If the number of I/Os
during this interval or the calculated average
CMR time for this path was not significant enough
for a meaningful analysis, '****' is displayed.
System action
The system continues processing.
Operator response
System programmer response
Refer to the corresponding reports for this Health
Check to determine which control units and channel
paths have delays. Use diagnostic tools (such
as the Resource Monitoring Facility (RMF™) or hardware diagnostics) to help determine
the source of the problem.
Problem determination
Source
Module
Routing code
Descriptor code
Automation
Reference Documentation
n/a