Explanation
VTAM® detected
that the session indicated is using 10 percent or greater of the bpBUF
buffer pool.
nodename1 and nodename2 are
the session partners for the session using the indicated percentage
of the pool. If VTAM does not
know a node name, the node ID is presented in the form subarea/element,
where subarea is the subarea portion of the network
address and element is the element portion of the
network address. Both subarea and element addresses
are given in decimal form.
bpBUF is the
name of the buffer pool.
percentage is the
percentage of this buffer pool used by this session.
If the
session between
nodename1 and
nodename2 is
using a large percentage of the buffer pool, one of the following
conditions probably exists:
- Either nodename1 or nodename2 is
malfunctioning. This could be a hardware, microcode, or application
program error that causes VTAM to
be flooded with data.
- Neither nodename1 nor nodename2 is
malfunctioning, but a large amount of data is being transmitted on
this session with no pacing in effect.
- A resource with many sub-resources is being activated or deactivated.
This requires a large concurrent number of I/O buffers.
- The maximum size of the I/O buffer pool has been defined too small
or the maximum amount of CSA that VTAM is
allowed to obtain is too small. These are the expansion limit parameters
on the IOBUF start option and the CSALIMIT start option.
System action
Message IST154I, IST1098I, or IST1099I is displayed
with this message.
- If message IST154I is displayed, the buffer pool is not expanded
at this time. When more storage becomes available, VTAM may try again to expand the buffer
pool. VTAM may be adversely
affected by this failure to obtain more buffers.
- If message IST1098I or IST1099I is displayed, processing continues.
- If the session is an SSCP-LU session, then the LU is deactivated,
and message IST1098I is displayed.
- If the session is an SSCP-PU session and the PU is a PU other
than an NCP, then the PU is deactivated and message IST1098I
is displayed. Since activation or deactivation of large NCPs
can cause situations where large number of I/O buffers are properly
in use nad NCPs are not known to cause HOT I/O situations,
NCP PUs will not be automatically activated.
- If the session is an LU-LU session (including CP-CP) then the
session is terminated, and message IST1099I is displayed.
Once VTAM has determined
that a session is using greater than 10 percent of the buffer
pool, a determination is made whether to automatically terminate
the session or deactivate the LU. If the percentage is greater than
or equal to the HOTIOTRM start option value, and the session type
is LU-LU, VTAM initiates termination
of all the sessions between
nodename1 and
nodename2. VTAM issues message IST1099I
when sessions are automatically terminated. If the percentage is greater
than or equal to the HOTIOTRM start option value, and the session
type is SSCP-LU or SSCP-PU (and the PU is not an NCP), VTAM initiates inactivation of
the LU or PU. VTAM issues
message IST1098I when automatic inactivation has been initiated.
Operator response
- If it appears that the problem is caused by a malfunctioning device
LU, try to deactivate the device using the VARY INACT command. In
extreme cases, you might have to physically disconnect or power off
the device.
- If it appears that the problem is caused by a malfunctioning PU
and that PU was not automatically deactivated (HOTIOTRM was
not specified), try to deactivate the device using the VARY
INACT command. In extreme cases, you might have to physically disconnect
or power off the device.
- If it appears that the problem is caused by an activation or deactivation
of an NCP or other PU, an attempt can be made to reactivate the PU
SCOPE=ONLY. Once the PU is active, each line can then be VARYed ACTive.
Activating resources in this order, will lesson the peak demand for
I/O buffers.
- If VTAM continues to issue
this message, enter the DISPLAY BFRUSE command. Save the system log
and request a dump for problem determination.
System programmer response
- Ensure that session pacing is in effect for the session using
the largest percentage of the buffer pool. The BIND request unit contains
the values used for each session. See the z/OS Communications Server: SNA Network Implementation
Guide for
more information about session pacing.
- If message IST154I is displayed before this message, and the session
between nodename1 and nodename2 is
not using a large percentage of the buffer pool, the size of the buffer
pool was probably underestimated.
- If message IST154I was issued, use the explanation of code in
that message to determine which buffer pool you need to modify.
- You might want to redefine your buffer pool or CSA start options.
If the start option cannot be modified using the MODIFY VTAMOPTS command,
you must modify the VTAM start
options file (ATCSTRxx) and restart VTAM to use the start option.
- If you want VTAM to automatically
terminate these sessions, specify the HOTIOTRM start option with a
value that is less than or equal to percentage.
This start option can be modified using the MODIFY VTAMOPTS command.
- For additional information, see:
Routing code
Descriptor code