Pulse for TM1 – Architecture

Starting with v6, Pulse is using the REST API allowing a single Pulse application server to monitor many TM1 servers.

For Pulse v6, all TM1 instances need to be running Planning Analytics v2.0.8 (TM1 v11.6) or later.

When working with multiple TM1 servers, you can choose between installing one Pulse 6 server that will monitor many TM1 servers or installing Pulse 6 on each server:

There are many advantages of a centralized Pulse server such as:

• Pulse configuration is managed in one location.
• All information about all TM1 instances are stored in the same database.
• Migration can happen between TM1 servers without having to download from one Pulse server then to upload the package to another.
• Upgrading will be quicker as only one Pulse server will need to be upgraded.

Pulse constantly tracks your TM1 model. Pulse 6 is composed of four services that update one internal database, one GIT repository and one Elasticsearch cluster:

Pulse has four services: one for the monitoring, one for the application service, one for the Elasticsearch service and one for the Kibana application:

• Pulse Monitor: Monitor your TM1 instances by auto-discovering them in the windows registry. It needs to be installed on each TM1 server to pick-up service stats (needs to be installed on each TM1 server).
• Pulse Application Server: This service manages most of the functionalities that you can find on the Web client (documentation, source control, reports, Pulse security, Workbook and Worksheet history). The Pulse application server connects to TM1 using the TM1 REST API.
• Pulse Elasticsearch: This service populates the built-in Elasticsearch database. The Elasticsearch data can then be consumed by Pulse Kibana.
• Pulse Kibana: This service manages the Pulse Explorer client to explore and visualize Pulse data. Pulse Explorer is powered by Kibana.

All the data that Pulse monitors are stored in three repositories:

• An internal database: By default Pulse installs a h2 database and store all the configuration and high-level information about your TM1 instances.
• A GIT repository: To store all changes from your TM1 objects.
• One Elasticsearch cluster: To store very detailed data that can then be retrieved by the Pulse Explorer.

All the data that Pulse monitors can be consumed by 3 clients:

• Pulse Thick Client: A Windows thick client that can connect to your Pulse server to monitor all instances, its signature feature is the license optimization report.
• Pulse Web Client: From the Web client, you can access most of the Pulse features such as Live Monitoring, Documentation, Migration, Change Tracking and Reports.
• Pulse Explorer: It is a new Web client powered by Kibana to explore and visualize Pulse data.

When installing Pulse, you can choose to split its services into two servers. The Pulse Explorer services (Pulse Elasticsearch and Pulse Kibana) can be installed on a separate server. Instead of using the default h2 database, Pulse can be configured to send the data to a Microsoft SQL Server.

Pulse for TM1 can be installed on-premise or on any clouds. It just can’t be installed on the IBM Planning Analytics cloud servers. To monitor TM1 instances running on the IBM Cloud, Pulse will need to be installed on another server and it will connect using the TM1 REST API.