Network Management

Orion agent fails to restart after reboot

Windows Agent is not starting after the server is rebooted. Orion agent will not start automatically. Agent showing down. Agent not responding. Startup type: Automatic (Delayed Start). Delayed start Agent service. SolarWinds Agent service.

First published date

2/7/2020 7:56 PM

Last published date

7/1/2026 9:29 PM

Overview

Orion Windows agent does not start automatically after scheduled reboots of the server. Sometimes the server start up may be slow and interfere with proper startup of the Orion agent. If agents are not starting as they should after a server reboot, perhaps a service startup type should be modified on the server. This article instructs how to make this change.

 

Scenario 2: Many agents disconnected after reboot (SQL / service startup timing)

Product section

Orion Platform

Cause

This can be caused when the servers start-up process is slow and ends up interfering with the Agent initialization process.  

Resolution

Try setting the agent service to delayed start.

  1. Remote desktop to the problematic Agent-managed server.
  2. Start > Run > Services.msc
  3. Right-click "SolarWinds Agent" > Properties.
  4. Startup type: Automatic (Delayed Start).
  5. Click Apply.
  6. Test startup of the agent service by rebooting.

Contact SolarWinds Support if this does not resolve the issue.

Scenario 2: Many agents disconnected after reboot (SQL / service startup timing)

Applies when:

  • After reboot / patching, 10–30+ agents show as disconnected / unable to connect.
  • Network tests succeed from the node to poller (Test‑NetConnection <poller> -Port 17778).
  • Manually restarting SolarWinds services on the poller (or the agent services) makes all affected agents reconnect.
  • This repeats after every maintenance window.

Likely cause

On the main polling engine, core SolarWinds services try to connect to SQL Server / SolarWindsOrion‑OSDB before SQL is fully online.
Initial connections fail (SQL timeout / “Cannot open database” errors), and Agent Management / SWIS do not fully recover, so agents cannot register until services are restarted.

What to check

On the poller / main platform server around reboot time:

  • AgentManagement.Service.log, Orion.InformationService.log, Core.BusinessLayer.log
    • Look for:
      • SQL timeouts
      • “Cannot open database ‘SolarWindsOrion‑OSDB’ requested by the login”
      • Pre‑login / SSL handshake timeouts to SQL

Fix (platform side)

On each affected poller:

  1. Open Services.msc.
  2. Set these services to Automatic (Delayed Start):
    • SolarWinds Information Service
    • SolarWinds Orion Module Engine
    • SolarWinds Agent Management Service
    • SolarWinds Licensing Service
  3. Ensure SQL Server for Orion is configured to start reliably and quickly on reboot (DBA to review startup time and recovery).
  4. Reboot during a maintenance window and confirm:
    • No new SQL startup errors in logs.
    • Agents reconnect automatically without manual service or agent restarts.

If only single / few nodes are affected, continue to use the agent‑side steps in this KB (set SolarWinds Agent to Automatic (Delayed Start) on those nodes, check AV / MachineKeys, etc.).