TrueSight Component

Accessing Monitoring Studio X through TSPS

Monitoring Studio X's Web interface is normally accessed through a URL like https://agent-hostname:3443. Once the Monitoring Studio X Component has been installed and configured in TSPS, the Monitoring Studio X interface to interact with an agent can also be accessed through TSPS with the below URL:

https://tsps-hostname:port/studio-tsps-web/?agent=agent-fqdn:3181

where:

• tsps-hostname:port is the hostname and port of TSPS
• agent-fqdn is the fully-qualified domain name of the PATROL Agent you want to interact with
• 3181 is the port of the PATROL Agent

From TSPS, you can therefore access Monitoring Studio X on all connected agents. No communication takes place directly between your browser and the Agent.

If your browser can connect directly to the Agent, Monitoring Studio X will show a green icon with a link that redirects to Monitoring Studio X served through its own HTTP server directly on the PATROL Agent directly. This generally provides a much snappier interface because response times are significantly shorter.

Augmented TrueSight

TrueSight's user interface is augmented in several ways to link its various elements to their corresponding views in Monitoring Studio X on the appropriate PATROL Agent. Monitoring Studio X additions to TrueSight are marked with a purple icon.

Tip: If the purple icon is not displayed, restart your browser or delete your browser cache.

REST API

Monitoring Studio X's REST API is fully documented and is typically accessed through a resource URL like in the example below:

$ curl -X GET "https://agent-hostname:3443/rest/agent/log"

When the Monitoring Studio X Component is installed on TSPS, the same REST API is also available through TSPS itself, with a resource URL like in the example below:

$ curl -X GET "https://tsps-hostname:port/tsws/10.0/api/studio/agent-fqdn/3181/agent/log"

where:

• tsps-hostname:port is the hostname and port of TSPS
• agent-fqdn is the fully-qualified domain name of the PATROL Agent you want to interact with
• 3181 is the port of the PATROL Agent

Behind the scene, the REST API query is forwarded in PSL to Monitoring Studio X on the PATROL Agent through the Integration Service.

Essentially, this means that you can interact with the REST API of many agents from one single entry point: TSPS REST API.

Authentication is performed like the documented authentication mechanism for the TSPS REST API, or with RSSO and cookies.

Fun Facts

The HTTP Server of Monitoring Studio X KM can be safely turned off on the agents. However, this will prevent you from connecting directly to the Agent.

You may experience slow responses from the Monitoring Studio X interface in TSPS because the data has a long way to go before reaching your screen: your browser => TSPS => Integration Service => PATROL Agent => perform the actual task => Integration Service => TSPS => your browser. On slow networks, you may experience a few seconds delay between your actions and their results in the interface.

Monitoring Studio X needs only to be installed on the PATROL Agents, it does not need to be running or loaded.

As Monitoring Studio X does not need to be running, it does not require Java on the PATROL Agent.

Issue with TrueSight Presentation Server 11.3.03 Upgrade

The upgrade of TrueSight Presentation Server to version 11.3.03 while Monitoring Studio X Component is already installed and configured in your environment will prevent the proper loading of the Monitoring Studio X Component in TrueSight:

• The TrueSight interface no longer shows links to Monitoring Studio X
• The Component page no longer lists Monitoring Studio X

The problem is that the upgrade script inappropriately adds entries in the /opt/bmc/TrueSightPServer/truesightpserver/conf/custom/csr.conf configuration file, that override the Monitoring Studio X Component's default entries in /opt/bmc/TrueSightPServer/truesightpserver/conf/services/csr.conf.

To fix the problem, edit the /opt/bmc/TrueSightPServer/truesightpserver/conf/services/csr.conf and search for these 2 lines:

vm.args.classpath.282="${truesight.home}/lib/studio-rest-gateway.jar"
vm.args.classpath.283="${truesight.home}/lib/studio-tsps-core.jar"

Note: The numbers may vary depending on your environment.

Replace the entry numbers (282 and 283 in the above example) with numbers that are not listed in either this file (./conf/services/csr.conf) and ./conf/custom/csr.conf. In a typical setup, replacing 282 and 283 with 287 and 288 will work:

vm.args.classpath.287="${truesight.home}/lib/studio-rest-gateway.jar"
vm.args.classpath.288="${truesight.home}/lib/studio-tsps-core.jar"

Then restart TrueSight Presentation Server with the below commands:

$ ./tssh server stop
$ ./tssh server start

Everything should work as expected after the restart and a refresh of the TrueSight interface in your browser.

Errors when trying to access a Monitor with a backslash in its object ID

Some KMs (notably BMC's Windows KM and Microsoft SQL Server KM) use backslashes in the object ID of certain classes. The URL to display these objects in Monitoring Studio X in your Web browser and through the REST API contains the path to the object to display and may therefore contain backslahes.

Apache Tomcat, the Web server and Application Server behind TrueSight Presentation Server, is configured by default to forbid the use of backslashes in URLs. As a consequence, when you try to access a monitored object that contains a backslash in Monitoring Studio X through TSPS, you will get an error message (HTTP/400 Bad Request).

To solve this problem, you need to configure Apache Tomcat to allow backslahes in URLs. To do so, edit the ./conf/custom/csr.conf file in your TSPS installation directory, and add the following lines:

# Allow backslashes
vm.args.system.501=-Dorg.apache.catalina.connector.CoyoteAdapter.ALLOW_BACKSLASH=true
vm.args.system.502=-Dorg.apache.tomcat.util.buf.UDecoder.ALLOW_ENCODED_SLASH=true

Note: If there is already a vm.args.system.501 or vm.args.system.502 entry, use the next numbers in sequence.

Then restart TrueSight Presentation Server with the below commands:

$ ./tssh server stop
$ ./tssh server start

HA Installation

In a clustered environment, the installation of the Monitoring Studio X Component must be performed on each TrueSight Presentation Server (TSPS) node. The configuration however needs to be done only on the active node.

The below procedure is therefore recommended:

1. Install the Component on the standby node (B), following the above installation procedure.

2. Wait for this TSPS node (B) to restart and be back in standby status. Do not proceed with the next step until TSPS is ready on the standby node (B).

3. Install the Component on the active node (A), with the same procedure.

   This will trigger a restart of TSPS, and potentially a failover of the cluster (where node B takes over and becomes the active node).

4. Wait for the cluster to be back online. This may take several minutes. Be patient.

5. Connect to the TSPS interface, using the cluster address, and configure the Component, following the regular procedure.

6. Optionally, trigger a failover of the cluster by running a tssh server start on the active node.

   Once the cluster is back online, verify that the Monitoring Studio X TrueSight Component is still properly listed as Connected in Administration > Components.

   This optional step simply verifies that the failover mechanism will work safely for Monitoring Studio X.

To determine the exact hostnames and the status of each node in your clustered TSPS, you can run the below command on either node:

[tsps@dev-tsps-ha-1 bin]$ ./tssh ha status

+--------------------------------------------------------------------------------------------+
|   BMC TrueSight Presentation Server - Command Line Interface 2020 version 11.3.04          |
|   Copyright 2014-2020. BMC Software, Inc. as an unpublished work.  All rights reserved.    |
+--------------------------------------------------------------------------------------------+


Real-time Stats details
==============================
Node State :> Active
Active since :> Dec 7, 2020 6:15:00 PM



Primary node details
------------------------------
TSPS FQDN :> dev-tsps-ha-1.internal.sentrysoftware.net
Cache replication port :> 7800
Control port :> 8800



Secondary node details
------------------------------
TSPS FQDN :> dev-tsps-ha-2.internal.sentrysoftware.net
Cache replication port :> 7800
Control port :> 8800



Load Balancer
------------------------------
FQDN :> dev-tsps-ha.internal.sentrysoftware.net

HA Upgrade

Similarly, the upgrade of the Monitoring Studio X TrueSight Component must be performed on each node:

1. Upgrade on the standby node with tssh componenttype update.

2. Wait for TSPS to restart and be back in standby status. Do not proceed with the next step until TSPS is ready on the standby node.

3. Upgrade on the active node with the same procedure.

   Note: This will trigger a restart of TSPS, and potentially a failover of the cluster. Wait

Upgrades do not require to re-configure the TrueSight Component.

HA Removal

The removal of the Monitoring Studio X TrueSight Component in a clustered environment is a bit trickier. Follow the below procedure carefully:

1. Delete the Monitoring Studio X Component through Administration > Components in the TrueSight interface.

   • If the component is immediately deleted, you can skip to the 4th step.
   • If the component remains in the Deleting state, proceed with the next step.

2. Restart the active node with tssh server start.

   This will trigger a failover of the cluster and will synchronize the TSPS configuration from the active node to the standby node.

3. Wait for the cluster to be back online. This may take several minutes. Again, be patient.

4. Refresh your browser, and verify in the Administration > Components interface that the Monitoring Studio X TrueSight Component is no longer listed.

During the removal procedure, you may experience a blank screen issue when you refresh your browser (the TrueSight interface doesn't load). This is caused by the intermediary state of the cluster. The restart of the active node should solve the issue.