2. USAGE GUIDELINES › 2.3 Data Analysis Guidelines › 2.3.4 Analyzing Uptime
2.3.4 Analyzing Uptime
The CA MICS Analyzer Option for CICS assumes that the CICS
system is available when system records from pre-CMF 3.1.1
and ASG-TMON are processed. It equates uptime to the
duration of the system records. Duration is calculated by
subtracting the start timestamp from the end timestamp. For
pre-CMF 3.1.1, the start and end timestamps are both
available in the data. For ASG-TMON, only the start
timestamp is available. The CA MICS Analyzer Option for CICS
uses the start timestamp of the next system record as the end
timestamp of the current record. For the last ASG-TMON
system record encountered for each region, the CA MICS
Analyzer Option for CICS calculates the end timestamp by
adding the start timestamp, the total dispatch time, and wait
time for the interval.
CMF 3.1.1 and higher do not produce system records. The CA
MICS Analyzer Option for CICS assumes that the CICS system is
available when SMF type 110, subtype 57 records are
processed. The subtype 57 records are written at user
specified intervals. If a record represents either the first
period after recording is initiated or the last period before
recording is terminated the CA MICS Analyzer Option for CICS
sets duration to the sum of quasi- reentrant dispatch and
wait time, otherwise it sets duration to the user specified
recording interval. The CA MICS Analyzer Option for CICS
equates uptime to the duration of the subtype 57 records.
For CMF, the CA MICS Analyzer Option for CICS validates the
calculated duration against an interval value that you define
at product installation time. If the calculated duration is
less than zero or 1.5 times greater than the expected
interval, then DURATION is set to the specified interval
value.
Validation is not performed for ASG-TMON, since its system
records do not have to be written at an interval basis.
DETECTING A CICS OUTAGE
No event recorded by the supported CICS monitors indicates
the beginning of CICS processing. Task numbers are not a
reliable means of determining a CICS warm/cold start, because
the CICS task counter rolls over at 100000 tasks executed.
Therefore, the CA MICS Analyzer Option for CICS uses the
absence of "up" time to infer the presence of "down" time.
Actual detection of the down condition is difficult.
One CA MICS CICS management report shows CICS outage time.
The outage shown on the report is downtime (absence of
uptime) in any hour for which data was found. Gaps of no
data in the histogram show periods in which no downtime was
found. Unfortunately, the gaps can therefore represent
periods of CICS uptime AND periods of no data.
DETECTING MONITOR DATA LOSS
The CICS monitor you are using may be turned off at any time
during the day. This is a theoretical safety valve to enable
a reduction of CICS overhead during periods of system stress
or logging device unavailability. This condition looks
exactly like a CICS system outage to the CA MICS Analyzer
Option for CICS, because there is no data to indicate any
type of CICS activity.
SYSTEM INTERVAL/HOUR SYNCHRONIZATION
Pre-CMF 3.1.1 produces periodic system records based on a
user-defined interval. Usually, the interval is 15 minutes.
CICS Statistics 3.1.1 and above produce periodic statistics
records on a user-defined interval basis. The default
interval is 3 hours, but you should consider changing the
interval to 1 hour. You may change the interval using CEMT
SET STATISTICS and the EXEC CICS SET STATISTICS command.
Pre-CMF 3.1.1 and CICS Statistics 3.1.1 and above use timer
facilities that key on time of day to trigger the production
of a system/statistics record for an interval. The initial
time of day (time the first record is cut) is calculated to
be on an interval boundary. In the 15-minute case, system
records are cut immediately after the hour and at 15, 30, and
45 minutes after the hour.
Because the CA MICS data elements HOUR and ZONE are both
derived from end timestamp, system records with different
starting and ending hours will always be summarized in the
ending hour. This causes a misapplication of summarized
system activity up to the interval of the system records.
For instance, CMF system records produced on a 15-minute
interval can be misapplied by up to 15 minutes. The
following will clarify this:
Data for the period 08:30:00.20 to 08:45:00.20 is used by the
CA MICS Analyzer Option for CICS to construct a record whose
hour identification is hour '8'. Data for the period
08:45:00.20 to 09:00:00.20 is used by the CA MICS Analyzer
Option for CICS to construct a record whose hour
identification is hour '9', even though the data mostly
describes events which occurred in the eighth hour of the
day.
You should consider this misapplication when you perform any
analysis with the CICS System Activity File (CICCSY).
REPORTS
The CA MICS Analyzer Option for CICS provides you with
management reports to analyze CICS uptime. The reports are
produced automatically by the CA MICS operational jobs DAILY,
WEEKLY, and MONTHLY. In addition, MICF inquiries are also
provided to report CICS uptime. The MICF inquiries are
executed on an ad hoc basis. See Chapter 3 of this guide for
detailed descriptions and samples of the following reports:
o Management Objective Reports
- Daily CICS Availability Report
- Weekly CICS Availability Report
- Monthly CICS Availability Report
o Graphic Inquiries
- CICCD3 - Daily CICS Availability Report
- CICCW3 - Weekly CICS Availability Report
- CICCM3 - Monthly CICS Availability Report
o Printer Graphic Inquiries
- CICPD3 - Daily CICS Availability Report
- CICPW3 - Weekly CICS Availability Report
- CICPM3 - Monthly CICS Availability Report
In addition to the above reports, you can code SAS programs
to analyze CICS uptime using the following data elements:
FILE ELEMENT DESCRIPTION
------ ------- -----------
CICCSY CSYUPTM CICS Availability Time
CICCSY DURATION Recording Interval Time
CICCSY STARTTS Start Timestamp
CICCSY ENDTS End Timestamp
