Frequently asked questions
Monitoring, alerts, runbooks, AI, security, and pricing. If your question isn't here, email hello@fltln.io.
Monitoring
How is this different from Pingdom or Better Uptime?
Those tools check external HTTP endpoints. Faultline also monitors internal processes (background jobs, queues, cron tasks) via event ingestion, has multi-level escalation policies, and runs automated remediation runbooks. It's not just uptime; it's incident response end to end.
What can Faultline monitor?
HTTP endpoints (status code, response time, and optionally TLS certificate expiry), TCP ports, DNS records, SSL certificates, Docker containers, AWS ECS services, and heartbeats for anything that runs on a schedule or in a loop, such as cron jobs and queue workers.
Does it work with Docker and ECS out of the box?
Yes. You can connect Docker containers and AWS ECS services directly from the dashboard. Faultline polls container health and can restart services through runbooks: after an admin approves by default, or automatically for runbooks you switch to Auto-run.
Can I monitor a service that has no public URL?
Yes, with a heartbeat. HTTP checks run from the public internet and refuse URLs that resolve to localhost or a private network address, so for an internal service you have it ping a unique Faultline URL instead, and Faultline alerts when the pings stop.
How do heartbeats work?
You create a heartbeat check with an interval (how often the job should ping) and a grace period (extra slack for slow runs). Your job pings its URL each time it succeeds. If no ping arrives within interval plus grace, the check fails and goes through the same incidents, escalation, and runbooks as any other monitor. The job needs no API key; the token in the URL is the credential.
Will one slow response page me?
No. Every check is retried before it counts as a failure, and by default a monitor turns DEGRADED after 2 failed checks in a row and DOWN, opening a CRITICAL incident, after 3. At a 60-second interval a real outage opens an incident in about 3 minutes, and a blip that clears within 2 minutes never pages anyone. Two good checks in a row resolve it. You can change the thresholds per service.
How often do checks run?
Every 60 seconds on Free, down to 30 seconds on Pro (15 with the interval add-on), 15 seconds on Growth, and 15 seconds on Enterprise, or down to 5 by contract.
Does Faultline have public status pages?
Yes. A status page shows the services you choose, with public display names, 90 days of daily uptime, and recent incident updates, and visitors can subscribe by email to hear when an incident opens and resolves. Faultline's own status page at status.fltln.io runs on Faultline.
Is Faultline an APM? Does it replace Datadog?
No. Faultline has no distributed tracing and no log aggregation. It is detection plus the incident loop: monitoring, paging, runbooks, and post-mortems. If you need full observability, Datadog or New Relic is the right tool.
Alerts and on-call
Where can alerts go?
Email, Slack (incoming webhook), PagerDuty (routing key), or a webhook, which is HMAC-signed when you set a secret. Alert rules pick the minimum severity, the services, and the channels.
Does Faultline do on-call schedules and escalation?
Yes. On-call schedules support daily and weekly rotations, member ordering, and manual overrides for vacation swaps, with DST-stable handoffs. On Pro and above, escalation policies keep paging an unacknowledged incident step by step until someone owns it, and stop as soon as anyone acknowledges.
What happens when many services fail at the same time?
Faultline correlates them. When a shared dependency takes down several services at once, the storm collapses into a single page with the affected services attached, instead of paging you once per service.
Can I move off Opsgenie before it shuts down?
Yes. Opsgenie shuts down on April 5, 2027. Faultline replaces its on-call schedules and escalation and adds the monitoring and heartbeats Opsgenie never had. The migration guide walks through moving everything over.
Runbooks and AI
Do runbooks run automatically?
Only the ones you switch to Auto-run. Every new runbook requires approval: it is offered as a recommended fix on the incident and runs only when an Admin or Owner approves it. An auto-run runbook that has already started 3 times in the past hour is not started again; on-call is paged instead.
How does Faultline know a runbook actually worked?
It checks the service 2 minutes after every run. If the service is OK or degraded, the run is marked verified. If it is still down, Faultline looks once more 2 minutes later, and if it is still down then, it pages on-call at CRITICAL. Automation never fails silently.
What does the AI do?
It summarizes incidents as they open, drafts blameless post-mortems, recommends a fix from your own runbooks, drafts new runbooks from a service's incident history, and turns a plain-English description into a configured monitor. You can also ask questions in the in-app copilot console or the Slack bot.
Can the AI change anything without my approval?
No. AI output is advisory. Recommended fixes can only pick from runbooks you wrote, AI-drafted runbooks always arrive set to Require approval, and running anything still needs a person to say yes.
Can I connect Claude or another AI agent?
Yes. Faultline runs a hosted MCP server at mcp.fltln.io/mcp. Create a workspace API key (Pro and above) and add the server to Claude Code, Claude Desktop, or any MCP client. Claude can check health and investigate incidents, and your client asks you before it runs a runbook.
What incident data is sent to the AI model?
Only what each feature needs, such as the incident timeline and recent check results. AI features use Anthropic's Claude API, and Faultline never sends your API keys, AWS credentials, Slack tokens, or passwords. Resolved incidents are summarized and embedded with Voyage AI so future diagnoses can cite similar past incidents from your own workspace.
Is there a limit on AI usage?
Yes, a monthly token cap: 100,000 on Free, 2,000,000 on Pro, 5,000,000 on Growth, and unlimited on Enterprise. It resets on the 1st of each month (UTC). At the cap, AI features pause until the reset or an upgrade; monitoring, alerting, escalation, and runbooks keep working.
Security and data
How is my data handled, and is Faultline SOC 2 certified?
Not yet: SOC 2 Type II is on the roadmap. Today every workspace's data is scoped to its tenant on every read and write, API keys are stored only as SHA-256 hashes, Faultline never holds long-lived AWS keys (it assumes a role you control, 15 minutes at a time), and AI features never receive your credentials. The full details are at fltln.io/docs/security.
What access does Faultline need to my AWS account?
An IAM role you create, which Faultline assumes for 15 minutes at a time. Start with the read-only ECS policy; add the remediation policy only if you want runbooks to restart or scale services. Each workspace gets its own ExternalId, so no other Faultline customer can use your role even if they learn its ARN.
Does Faultline need SSH access to my servers?
No. To restart a service or container on your own host, you run a small remediation agent that polls Faultline for work. Faultline never opens a connection to your machines and holds no SSH keys.
How long is my data kept?
Individual check results are kept 7 days on Free, 90 on Pro, 180 on Growth, and 365 on Enterprise. Hourly uptime rollups are kept for 1 year, and incidents, timelines, runbook runs, and the audit log for as long as your account is active.
Can I self-host?
Not yet. Faultline is a fully managed cloud service today, and self-hosting in your own VPC is on the roadmap for the Enterprise tier. Your check data stays in your organization and is never shared.
Pricing and billing
Is there a free plan?
Yes. Free covers 5 monitors, 2 team members, 60-second checks, and 7-day retention, with no credit card. Every plan includes the full real-time dashboard.
Can I try a paid plan first?
Yes. You can start a 14-day trial of Pro or Growth from the Billing page, once per workspace, with no card.
Is pricing per user?
No. Each plan is one flat monthly price. Pro is $49 a month for 50 monitors and 10 team members; Growth is $159 a month for 150 monitors and 25 team members.
What happens when I exceed my monitor limit?
You'll get a warning in the dashboard. On paid plans, overages are billed at the per-monitor rate for your tier. On Free, new monitors are blocked until you upgrade or remove existing ones.
Is there an annual discount?
Yes. Paying annually saves 15%: Pro is $492 a year and Growth is $1,620 a year.
Can I get a refund or cancel anytime?
Yes. Every paid plan has a 30-day money-back guarantee: email hello@fltln.io from the account owner's address within 30 days of your first paid charge. You can cancel anytime under Settings, Billing, and you keep your plan until the end of the period you paid for.
Getting started
How quickly can I get set up?
Most teams are monitoring their first service within 5 minutes. Add a URL, set a threshold, and you're live. Escalation policies and runbooks take a bit longer but are optional to start.
Do I need to install anything?
No. HTTP, TCP, DNS, SSL, and ECS monitors are set up from the dashboard, and Docker monitors connect to your Docker API host. The faultline CLI (npm install -g faultline-cli, Node.js 18 or later) is optional and handy for heartbeats and scripting, and a cron job can ping its heartbeat with plain curl.
Who builds Faultline?
Chris Collareta, an engineer who has carried the pager. You can reach him at hello@fltln.io.