.There are a wide variety of metrics available to help
you monitor the performance of your IBM Cognos 8 installation.
They are listed in the following table with the type of entry they
apply to, a description, and the associated change type, time scope,
and gathering time .
Some metrics are reset when the service restarts. You can also
reset some metrics manually .
At the system and server levels, the metrics include all associated dispatchers. At the dispatcher level, metrics include all associated services. For server groups, metrics are for all the dispatchers in the group.
You can use session metrics to monitor user sessions. This is
useful for monitoring system trends such as usage patterns by time
of day and day of week. Session metrics are also useful for understanding
the context of other metrics. For example, if the number of sessions
is extraordinarily high, it could account for the queue length metrics being
higher than normal.
| Metric | Entry | Description | Change Type | Time Scope | Gathering Time |
| Number of Sessions | System | The number of currently active user sessions | Gauge | Point in time | On demand |
| Number of sessions high watermark | System | The maximum number of active user sessions since the last reset | Gauge | Since reset | On change |
| Number of sessions low watermark | System | The minimum number of active user sessions since the last reset | Gauge | Since reset | On change |
You can use queue metrics to determine if the system is keeping up with demand. For example, if requests spend too long in a queue, you may not have enough resources to meet demand.
There are queue metrics for services that use queues: report service and report data service.
There are also queue metrics at the system level for the following things:
Job queue, which contains metrics related to the internal queue used by all event management services.
Task queue, which contains metrics related to the internal queue used by all monitor services. This queue contains tasks until they are successfully completed.
SMTP queue, which contains metrics related to the internal queue used by all delivery services. This queue contains emails until they are sent.
Some of the metrics available for these queue metric groups must
be enabled to be displayed. For more information, see .
| Metric | Entry | Description | Change Type | Time Scope | Gathering Time |
| Latency | System Server/Server group Service | The average amount of time that requests have spent in the queue (in seconds). | Gauge | Since reset | On change |
| Number of queue requests | System Server/Server group Service | The number of requests that have passed through the queue | Counter | Since reset | On change |
| Queue length | System Server/Server group Service | The number of items currently in the queue | Gauge | Point in time | On demand |
| Queue length high watermark | System Server/Server group Service | The maximum number of items in the queue since the last reset | Gauge | Since reset | On change |
| Queue length low watermark | System Server/Server group Service | The minimum number of items in the queue since the last reset | Gauge | Since reset | On change |
| Time in queue | System Server/Server group Service | The cumulative amount of time that requests have spent in the queue (in days, hours, minutes, and seconds) | Counter | Since reset | On change |
| Time in queue high watermark | System Server/Server group Service | The maximum length of time that a request waited in the queue (in days, hours, minutes, and seconds) | Gauge | Since reset | On change |
| Time in queue low watermark | System Server/Server group Service | The minimum length of time that a request waited in the queue (in days, hours, minutes, and seconds) | Gauge | Since reset | On change |
You can use JVM metrics to monitor the Java Virtual Machine and
associated heap size (the amount of memory that is currently in
use). For example, if a dispatcher has been running for a long time
and heap usage is high, you may want to restart it. The maximum
heap size metric tells you if you have allocated a suitable amount
of memory to the JVM based on the amount o hardware memory available.
The current heap size, in relation to the maximum heap size, lets you
know if available memory is being used. If current heap size is
close to the maximum heap size, you may want to adjust tuning settings to
reduce the load on a particular JVM. Keep in mind that current heap
size may vary widely depending on the current load on the system.
| Metric | Entry | Description | Change Type | Time Scope | Gathering Time |
| Current heap size (bytes) | Dispatcher | The current size of the JVM heap (in bytes) | Gauge | Point in time | On demand |
| Initially requested heap size (bytes) | Dispatcher | The initial amount of memory that the JVM requests from the operating system during startup (in bytes) | Gauge | Point in time | On demand |
| Maximum heap size (bytes) | Dispatcher | The maximum amount of memory that can be used by the JVM (in bytes) | Gauge | Point in time | On demand |
| Up time | System Server/Server group Dispatcher | The length of time that the JVM has been running (in days, hours, minutes, and seconds) At the system, server, and server group levels, this is the highest value from all associated dispatchers. | Counter | Point in time | On demand |
You can use request metrics to monitor volume of requests, operational status of services, response times, and processing times. General request metrics include data for all services and are a consolidation of metrics for all dispatchers. Request metrics specific to a service include only data for that service.
At the system, server, and server group levels, the metrics include data from all associated dispatchers. At the dispatcher level, metrics include all associated services.
| Metric | Entry | Description | Change Type | Time Scope | Gathering Time |
| Current time | Service | The current date and time used by the service to interpret time values. Used only if the service has no clock synchronization mechanism | Counter | Point in time | On demand |
| Last response time | System Server/Server group Dispatcher Service | Processing time for the most recent successful or failed request (in days, hours, minutes, and seconds) | Gauge | Point in time | On change |
| Number of failed requests | System Server/Server group Dispatcher Service | The number of service requests that failed (a fault was returned) | Counter | Since reset | On change |
| Number of processed requests | System Server/Server group Dispatcher Service | The number of processed requests | Counter | Since reset | On change |
| Number of successful requests | System Server/Server group Dispatcher Service | The number of service requests that succeeded (no fault was returned) | Counter | Since reset | On change |
| Percentage of failed requests | System Server/Server group Dispatcher Service | The percentage of processed requests that failed | Gauge | Since reset | On change |
| Percentage of successful requests | System Server/Server group Dispatcher Service | The percentage of processed requests that succeeded | Gauge | Since reset | On change |
| Response time high watermark | System Server/Server group Dispatcher Service | The maximum length of time taken to process a successful or failed request (in days, hours, minutes, and seconds) | Gauge | Since reset | On change |
| Response time low watermark | System Server/Server group Dispatcher Service | The minimum length of time taken to process a successful or failed request (in days, hours, minutes, and seconds) | Gauge | Since reset | On change |
| Seconds per successful request | System Server/Server group Dispatcher Service | The average length of time taken to process a successful request (in seconds) | Gauge | Since reset | On change |
| Service time | System Server/Server group Dispatcher Service | The time taken to process all requests (in days, hours, minutes, and seconds) | Counter | Since reset | On change |
| Service time failed requests | System Server/Server group Dispatcher Service | The time taken to process all failed service requests (in days, hours, minutes, and seconds) | Counter | Since reset | On change |
| Service time successful requests | System Server/Server group Dispatcher Service | The time taken to process all successful service requests (in days, hours, minutes, and seconds) | Counter | Since reset | On change |
| Successful requests per minute | System Server/Server group Dispatcher Service | The average number of successful requests processed in one minute | Gauge | Since reset | On change |
The report service and batch report service have the unique metrics listed in the following table, which you can use to monitor report service processes.
| Metric | Entry | Description | Change Type | Time Scope | Gathering Time |
| Number of processes | System Server/Server group Report service and Batch report service | The number of report service processes currently running | Gauge | Point in time | On demand |
| Number of processes high watermark | System Server/Server group Report service and Batch report service | For system, server, and server group, the total of all Number of processes high watermark metrics for all associated resources For services, the maximum number of report service processes that ran at any one time since the last reset | Gauge | Since reset | On change |
| Number of processes low watermark | System Server/Server group Report service and Batch report service | For system, server, and server group, the total of all Number of processes low watermark metrics for all associated resources For services, the minimum number of report service processes that ran at any one time since the last reset | Gauge | Since reset | On change |
