Create Azure support requests in Azure SRE Agent

Why filing Azure support requests is hard

Creating Azure support requests in Azure SRE Agent eliminates the manual work of filing effective tickets. Normally you select the right service category, describe the environment and symptoms, and attach evidence like logs and metrics. Your agent handles all of this automatically.

Most tickets arrive under-documented. The support engineer asks for information you already collected during troubleshooting, adding hours to resolution. Worse, a misrouted ticket (wrong service or problem classification) can delay triage by days.

How your agent creates Azure support requests

Your agent builds the support request from diagnostic evidence it already collected. No copy-pasting, no going through the Azure portal support wizard, no guessing which service category to pick.

Step 1: The agent offers to create a ticket

At the end of any troubleshooting session, your agent presents a summary and asks what you'd like to do:

DIAGNOSTIC SESSION COMPLETE

Summary of what I found:
- Symptom:           VPN tunnel drops every ~50 minutes
- Suspected cause:   PFS missing on IPsec rekey
- Confidence:        High
- Boundary:          Inside Azure

Is the information I provided sufficient, or would you like me to
create an Azure Support Request with the full analysis attached?

  1. Sufficient - just save the report
  2. Yes, create a Support Request
  3. Just give me the PDF report
  4. I'd like to ask follow-up questions first

You can also trigger this step at any time by asking in chat:

Create a support request for this VPN issue

File an Azure support ticket - the ExpressRoute circuit isn't learning routes

Step 2: The agent collects your contact details

If you choose to create a ticket, the agent asks for your information in a single block:

Step 3: The agent resolves the correct routing

The agent maps the root cause to the correct Azure service and problem classification. For example, a VPN rekey failure routes to VPN Gateway > Connectivity and Performance > Site-to-Site, not to a generic networking category.

The agent validates that the service and problem classification IDs match before submitting, catching the common "product and area path do not match" error before it happens.

Step 4: The agent builds the ticket

The support request description is structured, not free-form:

1. Symptom as you reported
2. Environment: subscription, resource group, source and destination resources
3. Diagnostic mode: Active or Passive, which tools were used
4. Hop-by-hop findings: every network hop checked with pass/fail status
5. Key evidence: effective routes, BGP status, circuit metrics, log excerpts
6. What was ruled out: confirmed-healthy components
7. What remains unknown: gaps for the support team to investigate
8. Suspected root cause: clearly labeled hypothesis
9. Steps already tried: what you did before escalating
10. PDF attachment reference: full diagnostic report attached

The agent generates a PDF from the complete diagnostic analysis (including the Mermaid network path diagram) and attaches it to the ticket.

Step 5: Confirmation

After creating the ticket, the agent confirms with the ticket ID and a direct link to manage it in the Azure portal.

Requirements

Before and after using Azure SRE Agent for support requests

Related content

• Diagnose with Azure observability
• Track incident value