Hi Everyone,
We are using Elastic Cloud (SaaS) with Fleet-managed Elastic Agents on Windows servers. Some of our agents are intermittently switching between Healthy and Unhealthy states, and the status is not stable.
In the System integration, we are seeing the following error related to the Winlog input:
“device is not ready”.
The System integration currently has both Winlog and Metrics inputs enabled.
We would like to understand:
- If one input (for example, Winlog) fails, will the entire Elastic Agent become Unhealthy?
- What does the “device is not ready” error indicate for Winlog input?
- How can we identify the exact root cause for these intermittent non-fatal errors?
Currently around 4–5 agents are showing non-fatal errors intermittently.
Environment:
- Elastic Cloud (SaaS)
- Fleet-managed Elastic Agents
- Windows servers
- System integration with Winlog + Metrics inputs
Any troubleshooting suggestions or recommended logs to check would be helpful.
Thanks.