Recurring Azure DevOps agent connectivity failures and severe UI latency

Question

We are experiencing recurring Azure DevOps connectivity issues that are affecting our production deployment process.

Our self hosted build agents have been running reliably for approximately one year. Since late August, they have started repeatedly losing their connection to Azure DevOps Services. The problem occurred on August 26, again on September 2, and is occurring again today, September 3.

Typical agent errors include:

Agent connection error: The HTTP request timed out after 00:01:00.
The connection will be retried until the agent reconnects.

and failed pipeline runs with:

##[error]We stopped hearing from agent WIN-EI1047AHDGR.
Verify the agent machine is running and has a healthy network connection.
Anything that terminates an agent process, starves it for CPU, or blocks its network access can cause this error.
Pool: build-agents

During the same periods, the Azure DevOps web interface is also extremely slow. This has happened repeatedly on separate days and is not limited to a single failed pipeline run.

The agent infrastructure and network setup had been stable for approximately one year before these incidents started. We are not aware of any corresponding infrastructure changes on our side.

Business impact:

Our Azure DevOps pipelines are used for production deployments. The recurring connectivity failures make our deployment path unreliable and create a serious risk if an emergency production hotfix needs to be deployed.

Please investigate Microsoft side telemetry for our Azure DevOps organization during the affected periods and determine whether there are:

• increased request latencies • dropped or interrupted agent connections • networking or Azure Front Door issues • throttling

• Azure Pipelines service degradation

• organization or region specific service issues

We would also like Microsoft to determine whether the connection failures occur within our network, at the Microsoft network edge, or within Azure DevOps Services.

Affected periods include at least:

August 26, 2026, approximately 08:33 to 09:43 UTC

September 2, 2026

September 3, 2026, currently occurring again

We can provide the relevant _diag\Agent_*.log files, pipeline run IDs and additional timestamps on request.

This is a recurring production availability issue and not an isolated build failure.