Navigation

FAQ: MongoDB Agent

This addresses common questions about the MongoDB Agent.

General

What is the MongoDB Agent?

The MongoDB Agent can help you manage your MongoDB instances. Today, you need to install multiple software Agents to manage your instances. When we release the MongoDB Agent, you can update to this single binary on any server. Each Agent can back up, monitor, and/or manage instances. The choice is yours!

If you use any of the existing three Agents, Ops Manager will prompt you to update to the new MongoDB Agent. We explain the different possible update paths later in this document.

When will you launch MongoDB Agent in Ops Manager?

Between May and June 2019 for Ops Manager 4.1, we plan to release the MongoDB Agent. As it does today, Ops Manager gives you the option to update Agents on your schedule. If you do not update to the MongoDB Agent, you can remain on the final version of the three separate Agents. These Agents will not receive bug fixes or new features. MongoDB encourages you to update as soon as possible after release. Once you choose to update, you must update all clusters in the project. Once updated, the Ops Manager interface changes to the new flow and content for the MongoDB Agent.

When must I switch to the MongoDB Agent?

The best time would be when we make it available in Ops Manager. You can remain on the three Agents without bug fixes or new features.

Can I update per cluster or per node?

No. Updates to MongoDB Agent must be done for all clusters in a project simultaneously.

Will the Ops Manager UI change?

Yes. In areas like the Agent(s) and Servers tabs, some wording will change but nothing dramatic.

Will the Ops Manager API support the MongoDB Agent?

Yes.

Can I programmatically upgrade to MongoDB Agent via the API?

Yes.

Will the new installer use a different filename?

No. It will remain mms-automation to make the update easier.

I have a PEM Key File and Password(s) set for my Automation, Monitoring or Backup Agents? How’s that going to work?

When you update to MongoDB Agent, these settings are preserved. If you have set a PEM Key File and Password in the Security > Authentication & TLS/SSL section of Ops Manager, it uses the current Automation Agent PEM key file and Password as the MongoDB Agent PEM Key File and Password.

Any existing Monitoring and Backup PEM Key File and password(s) are exported into the new custom configuration section for Monitoring and Backup on the MongoDB Agent page. A warning displays at the Configure Cloud Manager Agents section (Security > Authentication & TLS/SSL) that those values are being overridden with values from the custom configuration section.

I only use the Monitoring Agent, what does this mean for me?

Rest assured, you will be supported with MongoDB Agent. In general, the update flow proceeds as follows:

  1. When the MongoDB Agent is available, you see the usual banner notification that your Monitoring Agent is out of date.

  2. When you are ready to update, click Update in the banner. A workflow that guides you through the update process starts.

    1. First, you read a brief explanation of what MongoDB Agent is with a link to relevant documentation.

    2. Second, you can enter any custom Monitoring configuration options you require for your Ops Manager project.

    3. Third, you can download and install the new MongoDB Agent. Ops Manager verifies that it has been installed correctly on the list of known servers.

      Note

      Previous authentication methods continue to work and MongoDB authentication information are entered in the same locations in the User Interface. (Monitoring user settings under the Monitoring Settings menu option of the ellipsis h icon menu.)

  3. Once the MongoDB Agent is installed on a server, it enables the Monitoring where it existed as a standalone Agent before.

  4. The MongoDB Agent also puts the old Monitoring Agents into standby mode and no longer shows them in the UI.

  5. The old standalone Monitoring Agents can no longer monitor instances. You can stop and remove them when ready.

All done! From this point forward, the MongoDB Agent can update itself once you Confirm and Deploy a new version. You no longer need to download the binary each time a new version is available!

I only use the Monitoring and Backup Agents, what does this mean for me?

Rest assured, you will be supported with MongoDB Agent. In general, the update flow proceeds as follows:

  1. When the MongoDB Agent is available, you see the usual banner notification that your Monitoring and Backup Agents are out of date.

  2. When you are ready to update, click Update in the banner. A workflow that guides you through the update process starts.

    1. First, you read a brief explanation of what MongoDB Agent is with a link to relevant documentation.

    2. Second, you can enter any custom Monitoring and Backup configuration options you require for your Ops Manager project.

    3. Third, you can download and install the new MongoDB Agent. Ops Manager verifies that it has been installed correctly on the list of known servers.

      Note

      Previous authentication methods continue to work and MongoDB authentication information are entered in the same locations in the User Interface. (Monitoring user settings are found via the Monitoring Settings menu option in the ellipsis h icon menu. Backup user settings are found on the Backup screen under Options > ellipsis h icon > Edit Credentials.)

  3. Once the MongoDB Agent is installed on a server, it enables the Monitoring and Backup where they existed as a standalone Agents before.

  4. The MongoDB Agent puts the old Monitoring and Backup Agents into standby mode and no longer shows them in the UI.

The old standalone Monitoring Agents can no longer monitor instances. You can stop and remove them when ready.

All done! From this point forward, the MongoDB Agent can update itself once you Confirm and Deploy a new version. You no longer need to download the binary each time a new version is available!

My clusters are already managed using all 3 Agents, what will happen?

Step right onto Easy Street! The update flow proceeds as follows:

  1. When the MongoDB Agent is available, you see the usual banner notification that your Agents are out of date.
  2. When you are ready to update, click Update in the banner. A workflow that guides you through the update process starts.
    1. First, you read a brief explanation of what MongoDB Agent is with a link to relevant documentation.
    2. Second, you see a list your current servers and you can update all of them to MongoDB Agent with the click of a single button. As a part of this update process the MongoDB Agent:
      • Stops any Monitoring and Backup Agents.
      • Enables Monitoring and Backup on the servers where Monitoring and/or Backup were running as an Agent.
      • Removes the stopped Monitoring and Backup Agent binaries.
      • Unlocks the mms-monitoring-agent or mms-backup-agent users in your MongoDB instances so you can delete them, if desired. The MongoDB Agent uses the mms-automation user to connect to your instances.

All done! From this point forward, the MongoDB Agent works as a single process for Automation, Monitoring and Backup.

What about the three MongoDB Agent Users?

MongoDB is consolidating the three MongoDB users that the Automation Agent currently uses (mms-automation, mms-monitoring-agent, mms-backup-agent) into one user. This user, mms-automation, is granted all needed permissions to automate, monitor, and back up instances. MongoDB is not removing the old MongoDB users in case you use those users for something else. However, Ops Manager will unlock these users in its interface as part of the update so you can delete them.

I mostly manage my clusters with the Automation Agent but I have Standalone Backup and Monitoring Agents, what does this mean for me?

You get both update flows: one for anything standalone as explained in the previous sections:

I only use the Monitoring Agent, what does this mean for me? or

I only use the Monitoring and Backup Agents, what does this mean for me?

For those servers where Automation manages Monitoring/Backup Agents, you follow the My clusters are already managed using all 3 Agents, what will happen? section.