TPDC Manual (G06.24+)

Using the SPAM Performance Data Charts and
Reports
Tandem Performance Data Collector (TPDC) Manual—526357-001
5-31
Disc Response Time (DRTCHRT)
Disc Response Time (DRTCHRT)
This 80-column chart displays the average response time, in milliseconds, of each
logical disk volume. This response time is the sum of the average time that a request
message spends in the request queue (external to the disk process) and the average
time that a request spends within the disk process (both queued and in service).
The chart displays the component parts of this response time, including:
•
External queue time
•
Disk process CPU time
•
Physical disk time (compensated for unmirrored or mirrored drives with serial or
parallel writes)
•
CPU contention time
•
Other contention time, which includes:
°
The percentage of all requests that encounter record or file locks (the % Blkd
column), which shows application-level contention for popular locked records
or files
°
Relatively long-lived requests, such as CONTROL-27 (write notification)
requests on a file to which writes are done infrequently, or a TMP “permission
to delete audit trail” request
°
Contention for disk cache semaphores
°
Serial versus parallel writes on a mirrored volume (The TMF audit trail volume
can be an example of this.)
The sort order is chosen to match the Relative Disc Balance chart (RDBCHRT).
Primary CPU, Channel, and Controller are also shown for quick reference to the
Average Disc Utilization chart (ADUCHRT). A volume is flagged with an exclamation
point if it is mirrored but currently both drives are primaried through the same controller.
This chart uses the DGLOBAL and DISCINFO entities.
Parameters used are:
•
LID—specifies the Measure Load ID
•
EXCLUDE—specifies a value under which to exclude the number of requests per
second
•
SECTION—specifies the report section number
The chart is sorted in descending order of total queuing and service time.