1 – CLEAR
Clear the node-specific, in-memory cache of counters.
Syntax
COUNTERS CLEAR
Command Qualifiers Defaults
/ASSOCIATIONS /ASSOCIATIONS
/CHANNELS /CHANNELS
1.1 – Parameters
None.
1.2 – Description
To clear (zero) the counters in a node-specific, in-memory
cache, issue the COUNTERS CLEAR command on that particular node.
The command creates the node-specific, in-memory section of
association and channel counters if it does not already exist.
Then it zeros all fields in the in-memory section. Note that the
counters will be zeroed without first merging their values into
the cluster-wide database of channel counters. If a cluster-
wide, on-disk database does not already exist, a new one will be
created. Finally, the fields in the on-disk database for numbers
of stored messages, message recipients, and message volumes are
set based on the entries in the PMDF queue cache database.
Either the association counters, or channel counters, or both,
may be cleared. The default is to clear both association and
channel counters.
If you want to update the on-disk database with the old in-memory
values before clearing them, then you should issue a COUNTERS
SYNCHRONIZE command before issuing the COUNTERS CLEAR command.
1.3 – Command Qualifiers
1.3.1 /ASSOCIATIONS
/ASSOCIATIONS (default)
/NOASSOCIATIONS
This qualifier specifies whether to clear the in-memory cache of
association counters.
1.3.2 /CHANNELS
/CHANNELS (default)
/NOCHANNELS
This qualifier specifies whether to clear the in-memory cache of
channel counters.
2 – CRDB
Create a cluster-wide database of accumulated association and
channel counters.
Syntax
COUNTERS CRDB
Command Qualifiers Defaults
None. None.
2.1 – Parameters
None.
2.2 – Description
A new, cluster-wide database of channel counters can be created
with the COUNTERS CRDB command. The new database will have all
counters zeroed except for the count of messages stored in each
channel. Those counts will be determined by entries in the PMDF
queue cache database. In addition, if an in-memory section for
association and channel counters on this node does not already
exist, it will be created as well.
Once the on-disk, cluster-wide database exists, you may use
the COUNTERS SYNCHRONIZE command to merge the information from
the node-specific, in-memory cache of counters into the on-disk
database.
3 – SHOW
Display the contents of the cluster-wide database of channel
counters.
Syntax
COUNTERS SHOW [channel]
Command Qualifiers Defaults
/HEADER /HEADER
/OUTPUT=file-spec None
/SYNCHRONIZE /SYNCHRONIZE
/TIMEOUT=seconds /TIMEOUT=120
3.1 – Parameters
channel
Optional channel name indicating the channel(s) for which to show
counters. May contain wildcards.
3.2 – Description
The contents of the cluster-wide channel counter database may be
displayed with the COUNTERS SHOW command. By default, before the
counters are displayed, an implicit COUNTERS SYNCHRONIZE command
will be executed, to attempt to synchronize each node-specific
cache with the main cluster-wide database. Specify /NOSYNCHRONIZE
to merely display the current contents of the database without
first synchronizing the node-specific caches.
Note that SYSLCK privilege is required to perform the
synchronization step.
Note that the output of PMDF QM's COUNTERS SHOW command is
currently not as detailed as the output of the DCL level PMDF
COUNTERS/SHOW command.
3.3 – Command Qualifiers
3.3.1 /HEADER
/HEADER (default)
/NOHEADER
Controls whether or not a header line describing each column in
the table of counters is output.
3.3.2 /OUTPUT
/OUTPUT=file-spec
Direct the output to the specified file. By default the output
appears on your display.
3.3.3 /SYNCHRONIZE
/SYNCHRONIZE (default)
/NOSYNCHRONIZE
Before displaying the counters, attempt to synchronize each of
the node-specific caches with the cluster-wide database. Specify
/NOSYNCHRONIZE to skip this synchronization step.
3.3.4 /TIMEOUT
/TIMEOUT=seconds
By default, QM will wait upwards of 120 seconds for the node-
specific caches to be synchronized with the cluster-wide
database. Should the synchronization step not be completed before
the specified time period, then QM will stop waiting and proceed
to display the information from the database. You may specify a
different period of time to wait with the /TIMEOUT qualifier.
This qualifier has no effect when /NOSYNCHRONIZE is specified.
3.4 – Examples
To display the counters information for all TCP/IP channels,
use the command
qm.maint> COUNTERS SHOW *tcp_*
Channel Messages Recipients Blocks
------------------------ ---------- ---------- ----------
tcp_local
Received 33 41 95
Stored 0 0 0
Delivered 33 41 95
Submitted 1 1 3
tcp_internal
Received 632 758 1453
Stored 1 2 10
Delivered 631 756 1443
Submitted 3 6 12
qm.maint>
4 – SYNCHRONIZE
Synchronize each of the node-specific, in-memory caches of
channel counters with the cluster-wide database.
Syntax
COUNTERS SYNCHRONIZE
Command Qualifiers Defaults
/TIMEOUT=seconds /TIMEOUT=120
4.1 – Parameters
None.
4.2 – Description
To synchronize each of the node-specific, in-memory cache of
channel counters with the cluster-wide database, issue a COUNTERS
SYNCHRONIZE command. The command will not return control back
to you until either all the caches have been synchronized
or a "timeout" period has elapsed. Should the timeout period
elapse, then control will be returned to you. However, the
synchronization process will continue in the background. Use
the /TIMEOUT qualifier to adjust the timeout period which has a
default value of 120 seconds.
Note that SYSLCK privilege is required to use this command.
Note that the COUNTERS SYNCHRONIZE command signals each PMDF
counters synchronization process in the cluster to perform the
synchronization-there should be one such process on each node
running PMDF. Note that on each node, the synchronization can
only be performed if the PMDF counters synchronization process is
running on that node.
Assuming that the PMDF counters synchronization process is
running on each node, then for each node the node-specific, in-
memory cache will be created, if it does not already exist. If
the cluster-wide, on-disk database does not exist, it will be
created. The in-memory cache values will be used to update the
on-disk database, and then the on-disk database values for stored
messages, recipients, and volume will be set by scanning the PMDF
queue cache database.
4.3 – Command Qualifiers
4.3.1 /TIMEOUT
/TIMEOUT=seconds
By default, QM will wait upwards of 120 seconds for the node-
specific caches to be synchronized. Should the synchronizations
not be completed before the specified time period, QM will
return control to you prompting you for another command.
The synchronization process will, however, continue in the
background.
5 – TODAY
Display PMDF's count of the number of messages processed so far
today.
Syntax
COUNTERS TODAY
Command Qualifiers Defaults
None. None.
5.1 – Description
PMDF's count of the number of messages processed so far today may
be displayed with the COUNTERS TODAY command.
5.2 – Examples
This example illustrates displaying PMDF's count of the number
of messages processed so far today.
qm.maint> COUNTERS TODAY
4263 messages processed so far today
30000 messages per day are permitted by your license
qm.maint>