Monthly AutoSys Availability Statistics utility?

Document ID : KB000014531
Last Modified Date : 14/02/2018
Show Technical Document Details
Introduction:

We need to be able to determine the availability of our AutoSys instances and report on this to our management on a monthly basis. is there a utility that will do this?

Question:

We need to be able to determine the availability of our AutoSys instances and report on this to our management on a monthly basis. I understand that this reporting capability is built into CA7. We need to produce this information now for AutoSys. How can this be done in the short-term? What do you suggest? 

Answer:

You can use the autoaggr command.

It produces hourly,daily,weekly,monthly stats from the scheduler.  

 

From the Autosys Reference manual these are the stats collected:

=====================

Statistics Reported by the autoaggr Command

The autoaggr command generates reports on the aggregated job, alarm, and scheduler statistics.

Notes:

? The time column in the report represents the hourly time after the time hack. For example, if the reported time is 10:00:00, it represents the aggregated statistics generated from the activity that occurred from 10:00:00 to 10:59:59.

? You can also aggregate statistics using the Compliance Application, but this application collects only certain job run statistics. For more information about the Compliance Application, see the Windows Implementation Guide or the UNIX Implementation Guide.

The reports display job statistics that show the total number of jobs in STARTING, RUNNING, SUCCESS, BYPASS, FAILURE, TERMINATED, ON_HOLD, OFF_HOLD, ON_ICE, OFF_ICE, ON_NOEXEC, OFF_NOEXEC, INACTIVE, and RESTART status.

The reports display the following alarm statistics:

AVERAGE RESPONSE TIME

Displays the average time taken to respond to an alarm.

TOTAL ALARMS

Displays the total number of alarms irrespective of alarm status.

Note: The statistics displayed in the report includes the count of all the alarms that CA Workload Automation AE generates.

UNANSWERED

Displays the total number of alarms that are open. This does not include alarms that are acknowledged or closed.

JOB FAILURE

Displays the total number of JOBFAILURE alarms due to jobs that are in FAILURE or TERMINATED state.

START FAILURE

Displays the total number of STARTJOBFAIL alarms.

MAX RETRY

Displays the total number of MAX_RETRY alarms.

MAX RUNTIME

Displays the total number of MAXRUNALARM alarms.

MIN RUNTIME

Displays the total number of MINRUNALARM alarms.

DATABASE ROLLOVER

Displays the total number of DB_ROLLOVER alarms.

SCHEDULER FAILOVER

Displays the total number of EP_ROLLOVER alarms.

SCHEDULER SHUTDOWN

Displays the total number of EP_SHUTDOWN alarms.

The reports display the following scheduler statistics:

JOB RUNS

Displays the total number of job runs.

JOB FAILURES

Displays the total number of jobs that are in FAILURE or TERMINATED status.

JOB FORCE START

Displays the total number of jobs that were force started.

KILL JOB

Displays the total number of jobs that are killed using the KILLJOB event.

SERVICE DESK REQUESTS

Displays the total number of jobs for which service desk issues were opened.

TOTAL EVENTS

Displays the total number of events (including internal events) that CA Workload Automation AE generates.

AVERAGE LATENCY

Displays the average latency (in seconds) in processing events, which is the average processing time that is required for all the events. The latency for an event is calculated by subtracting the time the event was inserted into the database from the time it was picked up and processed by the scheduler. The average latency value is calculated by adding all the event latencies and then dividing the total latency by the number of events.

MAX LATENCY

Displays the highest latency (in seconds) for processed events within a given time range. The latency for an event is calculated by subtracting the time the event was inserted into the database from the time it was picked up and processed by the scheduler.

AVERAGE LAG TIME

Displays the average lag time (in seconds) in processing events, which is the average processing time that is required for all the events. The lag time for an event is calculated by subtracting the time the internal processing starts for the event from the time the scheduler completes processing it. Unlike average latency, the average lag time does not include the time that is required to fetch the event from the database. The average lag time value is calculated by adding all the event lag times and then dividing the total lag time by the number of events.

MAX LAG TIME

Displays the highest lag time (in seconds) for processed events within a given time range. The lag time for an event is calculated by subtracting the time the internal processing starts for the event from the time the scheduler completes processing it. Unlike max latency, the max lag time does not include the time that is required to fetch the event from the database.using the autoaggr command.

It produces hourly,daily,weekly,monthly stats from the scheduler.  From the manual:

 

Statistics Reported by the autoaggr Command

The autoaggr command generates reports on the aggregated job, alarm, and scheduler statistics.

Notes:

? The time column in the report represents the hourly time after the time hack. For example, if the reported time is 10:00:00, it represents the aggregated statistics generated from the activity that occurred from 10:00:00 to 10:59:59.

? You can also aggregate statistics using the Compliance Application, but this application collects only certain job run statistics. For more information about the Compliance Application, see the Windows Implementation Guide or the UNIX Implementation Guide.

The reports display job statistics that show the total number of jobs in STARTING, RUNNING, SUCCESS, BYPASS, FAILURE, TERMINATED, ON_HOLD, OFF_HOLD, ON_ICE, OFF_ICE, ON_NOEXEC, OFF_NOEXEC, INACTIVE, and RESTART status.

The reports display the following alarm statistics:

AVERAGE RESPONSE TIME

Displays the average time taken to respond to an alarm.

TOTAL ALARMS

Displays the total number of alarms irrespective of alarm status.

Note: The statistics displayed in the report includes the count of all the alarms that CA Workload Automation AE generates.

UNANSWERED

Displays the total number of alarms that are open. This does not include alarms that are acknowledged or closed.

JOB FAILURE

Displays the total number of JOBFAILURE alarms due to jobs that are in FAILURE or TERMINATED state.

START FAILURE

Displays the total number of STARTJOBFAIL alarms.

MAX RETRY

Displays the total number of MAX_RETRY alarms.

MAX RUNTIME

Displays the total number of MAXRUNALARM alarms.

MIN RUNTIME

Displays the total number of MINRUNALARM alarms.

DATABASE ROLLOVER

Displays the total number of DB_ROLLOVER alarms.

SCHEDULER FAILOVER

Displays the total number of EP_ROLLOVER alarms.

SCHEDULER SHUTDOWN

Displays the total number of EP_SHUTDOWN alarms.

The reports display the following scheduler statistics:

JOB RUNS

Displays the total number of job runs.

JOB FAILURES

Displays the total number of jobs that are in FAILURE or TERMINATED status.

JOB FORCE START

Displays the total number of jobs that were force started.

KILL JOB

Displays the total number of jobs that are killed using the KILLJOB event.

SERVICE DESK REQUESTS

Displays the total number of jobs for which service desk issues were opened.

TOTAL EVENTS

Displays the total number of events (including internal events) that CA Workload Automation AE generates.

AVERAGE LATENCY

Displays the average latency (in seconds) in processing events, which is the average processing time that is required for all the events. The latency for an event is calculated by subtracting the time the event was inserted into the database from the time it was picked up and processed by the scheduler. The average latency value is calculated by adding all the event latencies and then dividing the total latency by the number of events.

MAX LATENCY

Displays the highest latency (in seconds) for processed events within a given time range. The latency for an event is calculated by subtracting the time the event was inserted into the database from the time it was picked up and processed by the scheduler.

AVERAGE LAG TIME

==========================