Blog
Product notes, reliability research, and practical guides from the ServiceAlert team.
Troubleshooting guides
Match a failed workflow to the affected service, region or component, then check the evidence before deciding what to do.
Supabase connection errors: outage or connection limit?
Check Supabase connection limits, pooler modes, query timeouts and official incident scope to narrow down a database or API failure.
Read more →Claude not working: chat, API or Cowork problem?
Separate Claude chat errors, API rate limits and overload, and Cowork file-access problems. Check the affected product before assuming a wider outage.
Read more →Zscaler not connecting: enrollment errors, ZIA or ZPA?
Troubleshoot Zscaler enrollment and connection errors. Match the error, affected service and cloud before assuming a wider outage.
Read more →Does an AWS Health event affect your account or region?
Match an AWS Health notice to its service, region and affected resources. Separate public incidents from account events before declaring application downtime.
Read more →Azure DevOps pipeline waiting for an agent: outage or queue?
Check Azure Pipelines queue messages, parallel-job capacity, agent demands and approvals before treating a delayed build as a provider outage.
Read more →Cloudflare 502 or 504: origin problem or provider outage?
Use request evidence, origin logs and official incident scope to investigate Cloudflare gateway errors without assuming a global outage.
Read more →Coolify down? Check the dashboard, server and application
Separate a Coolify dashboard problem from a self-hosted server, proxy or application failure. Check useful evidence before treating it as a wider outage.
Read more →GitHub Actions jobs stuck in a queue: what to check
Separate an Actions outage from runner availability, concurrency and approval delays, then collect the evidence needed for the next step.
Read more →How to read a Jira outage update: scope and recovery
Use a real Jira incident to separate affected scope, recovery and resolution, then check whether your team's failed workflow matches the provider notice.
Read more →Research and product notes
Reliability findings, incident response practices and ServiceAlert product updates.
What the March 2026 AWS Middle East outage changed for resilience planning
AWS facilities in the UAE and Bahrain sustained physical damage during the March 2026 conflict. The incident is a useful test of assumptions about availability zones, regions, and recovery plans.
Read more →Monitoring certificate expiry and basic TLS health
Basic SSL monitoring catches certificate expiry, hostname mismatches, and served-chain failures before they become customer-facing outages.
Read more →Make incident alerting a team system
How to replace one-person alert ownership with shared routes, clear service owners, and a response path the team can test.
Read more →Monitoring AI API dependencies in production
A practical monitoring plan for products that depend on hosted AI models, including application metrics, provider notices, fallbacks, and incident ownership.
Read more →What to do during an AWS outage
A response checklist for confirming customer impact, identifying affected services, communicating clearly, and reviewing recovery options.
Read more →Planning monitoring across multiple cloud providers
How to inventory cloud and SaaS dependencies, classify their impact, assign ownership, and route useful alerts.
Read more →Routing outage alerts to Slack, Teams, and Discord
How to choose an alert channel, limit noise, protect webhook credentials, and make sure a vendor outage reaches an owner.
Read more →Running a useful incident review
A practical structure for documenting an outage, understanding contributing conditions, and turning the review into owned improvements.
Read more →How to map API and service dependencies
A practical process for finding external dependencies, assigning owners, recording customer impact, and keeping the map current.
Read more →Estimating the cost of cloud downtime
A worksheet for estimating lost transactions, engineering time, support work, SLA credits, and other costs tied to an outage.
Read more →Patterns in the 2025 cloud status archive
A careful reading of recurring patterns in vendor status notices, with links to the current reliability data and its limitations.
Read more →How to review the cloud incidents that affected your team in 2025
A method for turning vendor notices and internal incident records into a useful annual reliability review without relying on a generic biggest-outages list.
Read more →Ten reliability questions to ask a SaaS vendor
A procurement checklist covering status history, SLAs, incident communication, maintenance, dependencies, and recovery planning.
Read more →Reliability questions when choosing Slack or Microsoft Teams
A practical comparison of dependency scope, status information, emergency access, and backup communications for Slack and Microsoft Teams.
Read more →Why a vendor status page is only one part of outage detection
Vendor status pages provide confirmed incident context, but application telemetry and dependency monitoring are still needed to understand customer impact.
Read more →An incident response plan for third-party outages
A concise plan for identifying critical vendors, assigning owners, routing alerts, and responding when a third-party service fails.
Read more →Understanding SLA uptime percentages
How to convert an availability percentage into a time budget and why the contract's scope, exclusions, and measurement window still matter.
Read more →Current data
For current provider information, open service status. Completed monthly summaries are available in reliability reports.