Monitoring the Presentation Server environment
You can monitor the TrueSight Presentation Server environment, so that you are alerted when a service becomes unavailable.
Alerts are generated in the form of automated email or events, or both.Use the following table to understand the types of alerts generated for the monitored services:
Monitored service | Alerts generated | ||
---|---|---|---|
Events | Automated | Both | |
Remedy Single Sign-On server | ✅️ | ✅️ | ✅️ |
Database (PostgreSQL) service | ✅️ | ✅️ | ✅️ |
Index (Elasticsearch) server | ❌️ | ✅️ | ❌️ |
Failover of high-availability installations | ✅️ | ✅️ | ✅️ |
Components registered with the TrueSight Presentation Server (Only when disconnects) | ✅️ | ❌️ | ❌️ |
Database backup | ✅️ | ❌️ | ❌️ |
Self-monitoring email alerts
The monitored services are checked every 10 seconds for availability and an email is generated after five unsuccessful retries.
By default, self-monitoring email alerts are not configured. You can configure self-monitoring email from the command line.
To configure self-monitoring email alerts
- From a command prompt, go to the <Presentation Server installation directory>\TrueSightPServer\truesightpserver\bin directory.
Run the following commands to set up the email alerts:
Command
Value
Example
tssh properties set mailer.enabled
true to enable email alerts; false to disable email alerts.
tssh properties set mailer.enabled true
tssh properties set mail.to.addresses
The email address or a comma separated list of email addresses to send the alerts to.
tssh properties set mail.to.addresses user@xyz.com
tssh properties set mail.from.address
The "From" email address that you want to be displayed in the email alerts.
tssh properties set mail.from.address admin@xyz.com
tssh properties set mail.server.hostname
The email server host name.
tssh properties set mail.server.hostname mail.xyz.com
- Run the following command to apply the properties:
tssh properties reload - (Optional) Run the following command to view the properties that are set:
tssh properties list
Self-monitoring events
The Events page of the TrueSight console displays self-monitoring events for the following conditions:
- Unavailability of the Remedy Single Sign-On server
- Unavailability of the database (PostgreSQL) service
- Unavailability of components registered with the TrueSight Presentation Server
- Failover on high-availability-enabled installations
Self-monitoring events are indicated by theicon. The event severity determines the color of the icon.
Click an event timestamp to view details about the event.
Monitoring the Presentation Server PostgreSQL database performance
The TrueSight Presentation Serveruses PostgreSQL as the back end database for storing information about the components configured in it.
You can use the PATROL for PostgreSQL KM to monitor the information about TrueSight Presentation Server configured components.
The following table lists the recommended attributes to monitor:
Attribute | Description | Observed value |
---|---|---|
InstanceAvailability | Displays the PostgreSQL environment's configuration status. | 0-OK |
PGR_DB | Displays the status of the database availability. | 0-OK |
DBNumElementsWaiting | Displays the number of elements waiting to get a lock. | 0 (average) |
DBNumOfLockedElements | Displays the number of locked database elements. | 1 (average) |
DBCacheUsage | Displays the amount of cache memory usage in the PostgreSQL database. | 99.5% (average) |
SpaceUsedMB | Displays the amount of space used (MB) in the tablespace of PostgreSQL database. | 80MB space utilization for pg_default tablespace (average) |
Monitoring the Presentation Server Elasticsearch performance
The TrueSight Presentation Server stores events in Elasticsearch. You can use the PATROL for Elasticsearch KM to monitor the Elasticsearch database by using the standard process monitor KPIs.
Following are a few observations from the BMC lab testing:
- Approximately, 100K events were stored in the Elasticsearch database every day.
- Events are marked for deletion from the system under the following conditions:
- Closed events are marked for deletion 24 hours after the closure.
- Aged events are marked for deletion 30 days from the creation date.
- Internal processes in Elasticsearch clear the marked for deletion entries periodically.
- An average of 5 million documents are stored per day.
The number of documents that can be stored in Elasticsearch depends on the number of received events and event operations, such as closing of events.
The following table lists the recommended attributes to monitor:
Attribute | Description | Depends on | Observed value |
---|---|---|---|
Total number of documents across all nodes. | The number of events in the system and number of event operations | 5 million (average) | |
Number of documents deleted across all nodes. | The rate of closure and pruning of events | 1 million (average) | |
Number of items in the query cache across all nodes. | Depends on the number of event operations. | 9 million (approximately) | |
Rate of queries served in one millisecond (used as indicator to measure efficiency of queries). | Depends on the number of event operations. | 15 queries per millisecond (approximately) | |
Average response time of a query in milliseconds. | Depends on the number of event operations. | 0.2 milliseconds (approximately) |
Related topics
TrueSight Operations Management overviewMonitoring and managing events