--- title: "Failover protection with Agent Handlers best practice" slug: "failover-protection-with-agent-handlers-best-practice" tags: ["Product Usage", "ePolicy Orchestrator – On-prem", "ePO 5.10.x", "ePO 5.10.0"] updated: 2026-09-07T17:47:05Z published: 2026-09-07T17:47:05Z canonical: "docs.trellix.com/failover-protection-with-agent-handlers-best-practice" --- > ## Documentation Index > Fetch the complete documentation index at: https://docs.trellix.com/llms.txt > Use this file to discover all available pages before exploring further. # Failover protection with Agent Handlers best practice Agent Handlers allow any Trellix Agent to receive policy and task updates and report events and property changes if the ePO - On-prem server is unavailable. For example, an upgrade or network problem. Once multiple **Agent Handler** are deployed, they are available to agents as failover candidates. As long as the **Agent Handler** is connected to the database, it can continue serving agents. This includes any policy or task changes resulting from agent properties or from administrator changes before the ePO - On-prem server goes offline. The configuration file shared with the Trellix Agent contains a configurable fallback list of **Agent Handlers**. If needed, the Trellix Agent tries to connect through the list of **Agent Handlers** until the list ends or it can contact a valid, enabled **Agent Handler**. Failover between **Agent Handlers** is configured in one of two ways. ### Simple deployment failover In the simple deployment failover, two **Agent Handlers** can be deployed as primary and secondary. All agents initiate communications with the primary **Agent Handler**, and only use the secondary **Agent Handler** if the primary is unavailable. This deployment makes sense if the primary **Agent Handler** has better hardware, and can handle the whole load of the infrastructure. **Simple **Agent Handler** failover**![Simple Agent Handler failover](https://cdn.us.document360.io/0d453a50-e1da-4d4a-9409-fea7959614d8/Images/Documentation/uuid-2fa865ff-457b-7a1c-8931-8275cfe251e4.png?sv=2026-02-06&spr=https&st=2026-11-20T15%3A10%3A59Z&se=2026-11-20T16%3A10%3A59Z&sr=c&sp=r&sig=hmvRRX%2B5a1Fr9RmGrY8j4A9kptyJqJT7cvghO8e%2FDG0%3D) ### Failover with load balancing The second deployment combines failover with load balancing. Multiple **Agent Handlers** are configured into the same **Agent Handler** group. The ePO - On-prem server inserts each **Agent Handler** in the group into the list of **Agent Handlers** at the same order level. The Trellix Agent randomizes **Agent Handlers** at the same order level, which results in an equal load across all **Agent Handlers** in a particular group. **Failover with **Agent Handler** load balancing**![Failover with Agent Handler load balancing](https://cdn.us.document360.io/0d453a50-e1da-4d4a-9409-fea7959614d8/Images/Documentation/uuid-cd1cc014-3d9e-9815-94f2-b07ff0ded2a8.png?sv=2026-02-06&spr=https&st=2026-11-20T15%3A10%3A59Z&se=2026-11-20T16%3A10%3A59Z&sr=c&sp=r&sig=hmvRRX%2B5a1Fr9RmGrY8j4A9kptyJqJT7cvghO8e%2FDG0%3D) Agents failover between all **Agent Handlers** in a group before failing through to the next **Agent Handler** in the assignment list. Using **Agent Handler** groups results in both load balancing and failover benefits.