On-Demand Runbooks

An on-demand runbook enables you to execute a runbook manually and immediately, which is distinct from incident runbooks and lifecycle runbooks that execute in response to specific events.

On-demand runbooks are useful for cases such as:

• Discovery — List everything known about an entity from various products (Riverbed and third-party).

• Experimentation/testing — Test a runbook being defined for remediation to validate it before automating it.

• Routine IT troubleshooting — The IT desk received a report of an issue at a specific time; execute a runbook for the reporting user, location, or application at that time.

• Verification — An incident occurred yesterday and is now fixed. Execute a runbook to validate the fix and ensure that everything is back to normal.

• Change validation — A network configuration was performed; verify that everything is still up and running, based on a checklist of items to validate.

• Integration — Integrate with a third party vendor and send it specific network data collected from Riverbed products.

On-demand runbooks are listed on their own On-demand Runbooks page, separate from other runbook types.

Create an on-demand runbook by clicking New on the On-demand Runbooks page to open the Runbook Editor. The new runbook is started on the canvas with an Input node added by default.

On-demand runbook definitions are similar to other Runbook types, with these exceptions:

• An on-demand Runbook starts with an Input node, not a Trigger node. The Input node requires you to explicitly supply the details that give the on-demand runbook the context it needs to be able to execute, information that an incident Runbook acquires implicitly from its Trigger node. (This also means that Decision nodes used in on-demand runbooks don't have a Trigger category.) On-demand runbooks can use any of the following entity types as inputs:

  • None (Execute the runbook on all the data without specifying any input.)

  • Interface

  • Device

  • Application

  • Host

  • Client

  • Server

  • Location

  • Client Location

  • Server Location

• Since on-demand runbooks are executed manually at your initiation, they are not managed as automations like triggered runbooks (incident, lifecycle, and external) are.

• On-demand runbooks do not use/support Impact Statements.

• On-demand runbooks do not use/support incident variables.

When you create an on-demand runbook that specifies an entity type, you will be prompted to provide that entity at execution, and that entity then can be used to filter all the data queries within the runbook. For example, for an application input, you will be prompted to provide an application; if you then specify that you want to execute for the Exchange application, you can define the runbook such that all the runbook's Data Query nodes use Exchange as a filter in order to furnish additional information about Exchange.

Once the on-demand runbook has executed, its output is listed on the Runbook Analyses page.