Get support
Use Request Support in the deployed managed application or the contact shown in its approved Marketplace listing. A source-code portal URL or internal channel is not an approved customer support route unless it appears there.
Keep one support record from initial impact through customer-validated closure.
Submit through the managed application
- Open the managed application Overview.
- Select Request Support.
- Enter Your Email Address.
- Enter a concise Request Subject.
- In Request Description, state the impact, start time, affected users or services, exact symptom, and troubleshooting already completed.
- Review the form for credentials, personal data, and customer content.
- Submit the request and retain the resulting reference.
Classify the impact
Use the severity term defined in the approved Marketplace support agreement. If the agreement uses P1–P4, classify impact consistently:
| Severity | Customer impact example |
|---|---|
| P1 critical | Production service unavailable, widespread data ingestion stopped, or active security impact with no workaround |
| P2 high | Major production function degraded or a material group blocked with no acceptable workaround |
| P3 medium | Limited degradation, non-critical function affected, or an acceptable workaround exists |
| P4 low | Information request, documentation question, or planned change guidance |
This table helps describe impact; it does not create a response-time commitment. Support hours, initial-response targets, escalation intervals, and service exclusions come from the approved offer or service agreement.
Include
- managed application resource ID;
- subscription and application resource group names when approved for sharing;
- deployment region and plan;
- problem start time and time zone;
- exact user-visible symptom;
- affected endpoint or operation without credentials;
- whether one or all users and networks are affected;
- last known successful time;
- Azure deployment or correlation ID;
- redacted error code and message;
- diagnostic steps already completed;
- business impact and required response time.
Use this subject pattern:
[P1-P4] <application name> - <region> - <short symptom>
Use this description structure:
Impact:
Started:
Last known good:
Affected users, sources, or endpoints:
Observed error:
Expected result:
Recent approved change:
Checks completed:
Correlation or deployment ID:
Safe attachment list:
Contact and time zone:
Do not include
- passwords, one-time passwords, API keys, tokens, cookies, or headers;
- private keys or complete certificate bundles;
- unredacted ARM parameters or environment dumps;
- customer documents, indexed records, or personal information;
- credential-bearing Git remotes or internal automation endpoints.
If the issue is an Azure purchasing, subscription, policy, or platform-service problem rather than application behavior, use the Azure support path approved by your organization.
Escalation and communications
- Use the ticket reference for updates; do not split one incident across unapproved channels.
- State immediately when impact or scope increases.
- Ask for an escalation through the support route when the response target in the approved agreement is missed.
- Record maintenance, workaround, recovery, and closure decisions.
- For a suspected credential or data exposure, use the approved security incident route rather than attaching sensitive evidence to a normal ticket.