Root cause analysis when your stack lives on someone else's servers
When your dashboards light up, the first question is always "is it us?" Pingoru watches the official status page of every cloud and SaaS service your app depends on, so you can rule out — or confirm — a third-party outage in the time it takes to open a tab.
5 monitors free forever · No credit card
Real-shape feed of recent vendor incidents — when your app's p95 spiked at 14:05 UTC, the first row tells you AWS us-east-1 was already in trouble. RCA begins and ends with that one row.
How RCA actually works with Pingoru
A modern app talks to thirty or more services it doesn't own — AWS, Stripe, GitHub, OpenAI, Cloudflare, your CDN, your email and auth providers. When any one of them blips, your dashboards light up the same way they do when it's a bug in your own code. The hard part of RCA isn't noticing — it's figuring out which door the problem came through.
- 1 Customer reports a problem at 14:23 UTC. Open your incident feed and look for any vendor activity that overlaps with that window. The dashboard shows every monitor you've added, sorted by latest activity, with active incidents surfaced at the top.
- 2 Spot the upstream cause. AWS us-east-1 has a major-outage incident that started 14:18 UTC — five minutes before your customer noticed. The row shows duration, severity, and which specific service ("EC2 us-east-1", not just "AWS"). That's your RCA in one row.
- 3 Read the full vendor timeline. Click into the incident and you get every status update the vendor posted, with exact timestamps — Investigating → Identified → Monitoring → Resolved — and the vendor's own update body for each step. No need to load their status page, no need to scroll back through their history; the whole story is right there.
- 4 Export for the postmortem. Pull the incident (or the whole 90-day window) as CSV or JSON. Start time, end time, duration, every status update, affected services. Drop into your postmortem doc with one paste — no screenshotting, no copy-paste from twenty status pages.
- 5 Spot patterns over time. Pull 90 days of vendor incidents and overlay against your own alert log. The same vendor showing up three times this quarter is a vendor-renewal conversation; a recurring 4am maintenance window aligning with your peak hour is an architectural one.
Every status change, every update, every timestamp
Real outages don't go from "up" to "down" in one step. Vendors walk an incident through a lifecycle, posting an update at each stage:
- Investigating — vendor noticed something off; cause unknown.
- Identified — they know what's broken; mitigation in progress.
- Monitoring — fix deployed; watching for recovery.
- Resolved — confirmed back to normal.
- Scheduled / In progress / Completed — for planned maintenance.
Pingoru captures every one of those status changes with the vendor's exact timestamp and update body — not "AWS was down for 38 minutes" but "AWS was investigating at 14:18, identified the cause at 14:24, started monitoring at 14:42, resolved at 14:56."
Why it matters for RCA: when you're piecing together what happened, the difference between "investigating" and "identified" tells you when the vendor knew what was wrong — useful for both your own write-up and any SLA-credit conversation down the line.
Exports built for postmortems and SLA reports
Every incident in your dashboard can be exported as CSV or JSON for offline analysis. The export includes:
- Headline fields — title, started_at, resolved_at, duration, impact, affected services
- Full update history — every vendor status change with its timestamp and update body, in order
- Component-level detail — exactly which services flagged (e.g. "EC2 us-east-1", not just "AWS"), per-incident
- Provider context — provider name, slug, and canonical status page URL so the exported row is self-explanatory in a doc
Drop the export into your postmortem doc, your SLA-credit calc, or your vendor-renewal prep. No screenshotting, no copy-paste from twenty different vendor status pages. Free plan exports the last 30 days; Premium exports the full year of stored history.
Vendors we already watch
Amazon Web ServicesPlus AWS (every region × service), Google Cloud, Azure, Microsoft 365, Salesforce, Adobe Creative Cloud, PagerDuty, Twilio, Sentry, and thousands more in the full directory.
Skip the detective work
Instead of opening twenty status-page tabs at the start of every incident, look at one screen. Every vendor you watch, with current state and any open incidents — in one feed.
Down to the service
We don't just say "AWS is up." We say "EC2 us-east-1 is degraded; everything else is fine." That's the level of detail RCA actually needs — the service you specifically use, not a country-wide rollup.
Full update timeline
Every status change the vendor posted — Investigating → Identified → Monitoring → Resolved — with exact timestamps and the vendor's own update body for each step. No need to chase the vendor's status page.
CSV + JSON exports
Pull any incident, or a full date range, as CSV or JSON. Standard fields plus the complete update history and per- component impact. Drop straight into your postmortem doc or SLA-credit calculation.
Permanent history
Vendors typically expose 30 days of past incidents publicly. Once we've seen one, it lives in your account forever — so a year from now you'll have a year of vendor history per provider, even after it rolled off the vendor's own page.
Routed to where you incident
Pipe alerts to your incident channel — Slack, Discord, Teams, PagerDuty, or a signed webhook. The first message in a war-room thread becomes "AWS just opened an incident matching ours" instead of "anyone else seeing this?"
Who runs RCA with Pingoru?
- SRE and platform engineers — fastest possible "is it us or them?" answer at the start of an incident, before you commit to debugging your own code for an hour.
- Engineering managers — postmortem prep that doesn't require chasing down vendor screenshots a week after the fact.
- Customer-success and support teams — confirm a vendor is actually down before replying "we're investigating" to twenty customer tickets in a row.
- On-call rotations — middle-of-the-night pages where the first 30 seconds of "is it us?" is the difference between going back to sleep and a four-hour debugging session.
- Indie founders — one-person ops teams who can't keep tabs on every vendor manually but still need to answer customers fast when something breaks.
Frequently asked
What is root cause analysis?
Root cause analysis (RCA) is the process of figuring out WHY something broke — not just what's failing right now. For apps that depend on dozens of third-party services, the answer is usually one of three things: a bug in our code, a problem with our infrastructure, or a vendor we depend on is having an issue. The third bucket is the one Pingoru exists to confirm or rule out fast.
How is Pingoru different from APM tools (Datadog, New Relic, Sentry)?
APM tools tell you WHAT broke — p95 latency spiked, error rate climbed, a specific endpoint started returning 500s. They don't tell you WHY when the answer is 'AWS us-east-1 is having a bad day.' Pingoru fills that gap: a real-time view of every vendor's official status page so the first thing you do during an incident is open Pingoru and check the timestamp against active vendor incidents. Most teams use both together — APM for the symptom, Pingoru for the cause.
How fast can I confirm a vendor outage during an incident?
About as fast as you can open one tab. The dashboard shows every vendor you've added with their current state. If something's red, you have your answer. If everything's green, the issue is on your side and you can stop the vendor-status detective work and focus on actual debugging.
Do you capture the full incident timeline, or just the headline?
Full timeline. Every status update the vendor posts — Investigating, Identified, Monitoring, Resolved (or Scheduled / In Progress / Completed for maintenance) — is stored with its exact timestamp and the vendor's update body. So when you're writing a postmortem you can cite the moment the vendor first knew something was wrong, the moment they identified the cause, and the moment they confirmed it was over — not just an opaque 'AWS was down for 38 minutes.'
Can I export incident data for postmortems or SLA reports?
Yes — every incident exports to CSV or JSON, including the full update timeline (status + body + timestamp per update), affected services at the component level, and provider context (name, slug, canonical status URL). Export a single incident or a full date range. Free plan exports the last 30 days; Premium exports the full year of stored history. Drop straight into your postmortem doc, vendor-renewal prep, or SLA-credit calc — no screenshotting from twenty status pages.
Can I cross-reference incident timestamps with my own alert log?
Yes. Every incident is stored with its exact start time, end time, and the affected services. When you're writing a postmortem, you can scroll the relevant 24-hour window and overlay your team's own timeline (deploy log, alert log, customer reports) against any vendor incidents that overlap.
Does it cover the vendors my stack depends on?
6,116+ providers tracked, including every AWS region × service combo, every GCP product, every Azure service, Stripe, GitHub, OpenAI, Anthropic, Cloudflare, Slack, Twilio, Salesforce, Microsoft 365, Google Workspace, and the long tail of smaller SaaS tools. If you depend on it, we probably already watch it. Anything missing? Request it and we'll usually add it within 24 hours.
How far back does the incident history go?
Vendors typically expose 30 days of recent incidents through their public status pages, which is what Pingoru ingests. From the moment we add a vendor to our catalog, we keep every incident permanently — so a year from now you'll have a year of history per provider, even for incidents that have rolled off the vendor's own page.
Is there a free tier for small teams doing RCA?
Yes. 5 monitors free forever, email + webhook alerts, 30 days of incident history, and full access to the public provider pages so anyone on your team can pull up vendor history without an account. Premium ($15/mo) bumps that to 50 monitors, 1 year of history, 10 team seats, and unlimited notifications.
Related: cloud status monitoring (the buyer-intent framing), SaaS monitoring (business-tools framing), incident timeline (the dashboard view), notifications (where alerts go).
Cut RCA from hours to seconds.
Free forever for 5 monitors. $15/mo for 50 + team features. No credit card.
Start monitoring free →