New AI Incident Response, Multi-Region Agents, and Custom-Domain Status Pages — May 2026
Services Pricing Dashboard

Changelog

What's new at ServiceAlert.ai

Get notified when we ship new features and improvements.

May 2026

Slack-native AI SRE agent — investigate, hypothesize, fix

A real teammate in your incident channel. Mention @servicealert investigate when an alert fires and within seconds the agent posts a threaded investigation: probable root-cause hypotheses ranked by confidence, the actual log lines that triggered the call, the most-recent deploy that touched the affected service, the related declared incidents in the last 14 days, and a one-click code-pointer link to the suspect file in your GitHub repo. Slack reactions act as actions: a thumbs-up promotes the leading hypothesis to the channel, an X silences a flapping monitor with a stated reason, and a siren promotes the alert to a full declared incident with all the responder routing that comes with it. Confidence is calibrated weekly against incident outcomes so the agent gets sharper over time, not just chattier.

Open Incident Management →
May 2026

Multi-region monitoring agents you install yourself

Run uptime checks from anywhere, your AWS regions, your on-prem datacenter, your customer's VPC, a branch office. A single Go binary registers itself with your account, joins the monitor pool, and shows up as another region next to ServiceAlert's central probes. Per-monitor alert quorum (any / majority / all) means a single-region flake on one carrier no longer pages you at 3am. Region pills on every monitor show last-check latency from each source. Hairpin-NAT guard skips checks where a target resolves to the agent itself so private-network monitors stay accurate. The agent ships as a systemd unit with proper Linux capabilities so ICMP and TCP probes run without root, and the register-agent CLI writes the env file at 0600 so the registration token never lands in shell history.

View Monitoring Agents →
May 2026

Native on-call scheduling, with iCal feeds and cover requests

Real on-call rotations land natively in the dashboard. Build a schedule, set the rotation cadence (daily, weekly, custom), pin shifts to a timezone, and the rotation timeline renders the next 30 days color-coded by responder. Each schedule exports an iCal subscribe URL so on-call shows up automatically in Google Calendar, iOS Calendar, or Outlook. Cover-request workflow lets a teammate flag an upcoming shift and a covering responder accepts in one click, with an email trail for both. When an incident is declared, the commander field auto-fills from whoever is currently on-call. Notification emails for upcoming shifts go out via Postmark with a friendly From-name. Replaces PagerDuty schedules for most teams; PagerDuty itself still works as an integration for orgs that need its escalation engine.

Set up On-Call Schedules →
May 2026

Bring-your-own-domain status pages with automatic TLS

Customers can now serve their public status page at their own hostname, status.acme.com. Two-step DNS setup: first a TXT record proves ownership (verified via Cloudflare DNS-over-HTTPS, with the observed records surfaced in the dashboard so typos are easy to spot), then a CNAME points the hostname at our edge. Within ~5 minutes of the CNAME taking effect, our cert-issuance worker runs Let’s Encrypt HTTP-01, writes a per-domain nginx server block, reloads, and the page is live with a valid TLS cert. Renewals are automatic. Atlassian StatusPage charges $79/mo extra for this; we don't.

Connect a Custom Domain →
May 2026

Browser-step transaction monitoring with Playwright

The transaction monitor now runs a full Chrome browser when you need it, not just curl. Real JavaScript, screenshots at every step, authenticated cookies, and per-step assertions on text, status, or response body. The browser worker is a long-running daemon that drives a system Chrome over a unix socket, so memory stays bounded and step latencies are consistent across runs. Use it for multi-step user flows your synthetic tests can't fake (login → cart → checkout → success). Falls back to the lightweight HTTP transaction runner when a flow doesn't need a real browser.

Configure Transaction Monitors →
May 2026

BGP-level outage detection on the public outage map

The public /outage-map now has a third ledger next to vendor-declared and ServiceAlert-detected: BGP anomalies. A nightly RIPE RIS poller takes country-level prefix-count snapshots for 50+ countries and flags any country whose prefix count drops more than 5% below its 7-day trailing median. Catches Telegram in Pakistan, Twitter in Turkey, and ISP-scale events that vendors will never acknowledge. Each anomaly card links straight to its source data, opens a chart of the country's prefix count over time, and stays color-coded with an amber dot so it's immediately distinguishable from the other ledgers.

View the Outage Map →
May 2026

Real User Monitoring — network path, CDN PoPs, and choropleth

RUM dashboards now answer "is it the user’s network or our origin?" without a developer in the loop. Network Path splits TTFB into DNS, TCP, TLS, server processing, and download stages so a slow site stops looking like a single mystery number. CDN PoP attribution parses Server-Timing and cf-ray headers from real user requests to pin slowness to specific edge locations ("Cloudflare LHR is showing 4× normal LCP for 12 minutes"). The country choropleth ranks p75 LCP by user country, with web-vitals thresholds shading the map green / amber / red. Connection-type segmentation (4G / 3G / 2G / slow-2G / unknown) sits on every web-vital tile so a 9.2s INP on slow-2G doesn’t panic the team when the 4G p75 is fine.

Open the RUM dashboards →
May 2026

Outage event permalinks with custom social-card images

Every event on the public outage map is now a shareable permalink. The card’s share button opens a popover with Copy link, Open permalink, X / LinkedIn share intents, and the OS share sheet on mobile. When that link gets posted in Slack, X, or LinkedIn, the unfurled card is generated server-side per event: a 1200×630 PNG with the affected entity name, severity badge, current status, duration, and category. Cache key includes the event’s last-updated timestamp so the social card reflects the latest state. Catches journalists, ops engineers, and curious customers who would otherwise just screenshot the page.

Browse the Outage Map →
May 2026

Services catalog — group monitors under what they actually do

A new layer above monitors. Define your services (Checkout, Auth, API, CDN, Marketing) once, attach the monitors and certs that make them up, and the public status page rolls them up as components instead of a flat list of hostname checks. A new app_class tag on each monitor lets even un-grouped monitors land under the right service heading on a status page. Declared incidents can now link to a service so the recent-incidents panel on the service detail page populates automatically. On-call rotations attach to services so the right person gets paged when the right thing breaks.

Open the Services Catalog →
May 2026

mTLS, OAuth2, JSON Query monitors, and one-click cloning

HTTP monitors now support mTLS client certificates and OAuth2 client-credentials authentication, with secrets encrypted at rest using AES-256-GCM and protected against accidental exposure when the edit form is reopened (placeholder bullets round-trip safely). A new JSON Query monitor type fetches a JSON endpoint and asserts on a JSONPath expression, so an API health check can read $.healthy == true without a custom script. Every monitor row in the dashboard now has a one-click clone action that copies all settings (auth, regions, latency thresholds, service grouping) into a new monitor, ready to retarget at a different host or environment.

Open Monitors →
May 2026

Latency-degradation alerting and DNS propagation diff

Two new alert types beyond up/down. Latency degradation lets you set a per-monitor SLO with a multiplier on the rolling 7-day p95 baseline (default: 2×); when the live p95 trends past it, you get a degraded-performance alert instead of having to wait for the monitor to actually time out. DNS propagation runs your DNS monitor against five public resolvers in parallel (1.1.1.1, 8.8.8.8, 9.9.9.9, 208.67.222.222, and your domain’s authoritative) and surfaces a diff so you can see propagation lag directly. DNSSEC validation flags broken signatures.

Configure Monitors →
May 2026

Incident artifacts — screenshot, MTR, response body on every alert

When an HTTP monitor goes down, ServiceAlert now captures a packet of forensic evidence on the way out: a Playwright screenshot of the failing URL, an mtr network-path trace from the alerting region to the target, and the failing response body (truncated to 4KB). All three attach to the declared incident, render in the incident detail panel as a per-incident diagnostics expander, and are surfaced in the Slack AI SRE agent’s investigation thread as evidence. Eliminates the "by the time we looked, the site was back" problem on intermittent failures.

View Incident Artifacts →
May 2026

Per-component email subscriptions on public status pages

Subscribers no longer have to take the firehose. Every public status page now lists components with checkboxes; pick the ones you care about, drop in your email, and you only get notified when a service in that pick list changes status. Confirmation flow is double-opt-in with a Postmark-delivered link, the unsubscribe footer takes one click, and IPs aren’t stored in plaintext (salted hash, daily rotation). RSS and Atom feeds for the same per-component slices are auto-discovered via <link> tags so feed readers find them without configuration.

Configure Status Pages →
May 2026

Auto-silence detector for flapping monitors

Noisy monitors stop paging you on their own. A rules-based detector watches the per-monitor flip-flop rate and the alert-fatigue contribution per channel, and when a monitor crosses thresholds it gets auto-silenced with a stated reason ("flapped 8 times in last hour") that’s visible right on the monitor row. A Wake button in the dashboard, in Slack alerts, and via the new MCP server clears the silence in one click. The silence reason is captured for audit and feeds the AI SRE agent’s confidence model so future investigations weight noisy monitors lower.

View Monitors →
May 2026

MCP server — talk to ServiceAlert from any AI tool

A JSON-RPC MCP endpoint at /api/mcp exposes ServiceAlert as a first-class tool to Claude Code, Cursor, ChatGPT, and any other Model Context Protocol client. Read tools include monitor status, recent incidents, and uptime stats; write tools include silence_monitor, wake_monitor, and declare_incident with severity. Each MCP key has its own per-key rate limit and audit log; rotation is one click in the dashboard. Now you can ask "what monitors are down right now?" or "silence the API monitor for an hour" directly from your editor and the call lands authenticated, scoped, and recorded.

Read the MCP guide →
May 2026

Region landing pages for the 2300-service catalog

Four new SEO landing pages slice the catalog by geography: /catalog/north-america, /catalog/europe, /catalog/asia-pacific, /catalog/latin-america. Each lists region-relevant services (region-native providers like GoCardless, Klarna, Razorpay, Mercado Pago, dLocal, plus the global SaaS most teams in that region rely on) grouped by category. Per-region OG images for unfurls in social. The status-page graph is now fully connected, region pages cross-link into /category/<slug> pages, which cross-link into /status/<slug> service pages, which cross-link back to the regions a service is most often used in.

Browse North America services →
April 2026

Privacy-friendly analytics + Web Vitals, included

A single 3 KB snippet now captures pageviews, custom events, and Core Web Vitals (LCP, INP, CLS, TTFB, FCP) in one tag. The dashboard ships realtime active visitors, bounce rate, average visit duration, top pages and referrers, country, browser, OS, device, UTM attribution, and a Today / 24h / 7d / 30d bar chart. Privacy posture matches Plausible and Fathom: no cookies, no fingerprinting, no IPs stored. Existing Umami snippets work transparently, swap in the new site key and beacons start landing without any markup change. One more SaaS line item replaced.

Open the analytics dashboard →
April 2026

Live detection-lead transparency on the homepage

Every time ServiceAlert catches a vendor outage before that vendor's own status page acknowledges it, we now aggregate the result into a live counter on the homepage: total outages caught in the last 14 days, how many the vendor never acknowledged at all, and the median lead time on the rest. Drawn straight from our caught-first pipeline and refreshed automatically, so the number you see is the number we're actually posting against. The companion /accuracy page has the full log with an A-F transparency grade for each vendor. Our belief is that the vendors you depend on should be held to a standard, and we're making ours public first.

See vendor transparency grades →
April 2026

Setup guides, FAQ, and REST API reference are now public

The Teams & Monitors setup guide, the FAQ, and the REST API reference previously lived inside the authed dashboard, which meant you had to be a signed-in customer just to share a link to them. All three are now free to read without an account. Bookmark the API docs in your IDE, or drop the Teams setup guide into your onboarding doc so new hires can self-serve. The Teams guide in particular has every permission, alert-routing, and shared-monitor detail worth knowing, previously locked behind a login.

Read the Teams & Monitors guide →
April 2026

Dashboard refresh, faster load and a single settings surface

The authed dashboard got a visible speed bump and visual cleanup. A long-retired monitored-services grid that had been living below the modern cards is gone, and a handful of settings (Service Selection, Team Management, layout preferences) that had drifted into duplicated copies across two different pages are now consolidated into Settings, reachable in one click from the dashboard side panel. Nothing about your data or preferences changed, only the surface you interact with. Customers on slower connections will see the most benefit.

Open the dashboard →
April 2026

Vendor transparency warnings on per-service status pages

Status pages now surface a prominent callout when the monitored vendor's own status page is consistently late to acknowledge incidents. Every vendor in the catalog is graded A-F by ServiceAlert based on the average delay between detection (via official feeds + crowd reports + social signals) and the vendor's own acknowledgement. The warning only renders when the grade is D or F, honest vendors (A/B/C) get no callout because there's no story to tell. When it does appear, the message reads like "1Password's official status page is unreliable, on average 20.1 hours late to acknowledge incidents", with a link to the full /accuracy methodology and leaderboard. Replaces the previous "STATUS PAGE ACCURACY: D" stat tile that was buried in the per-service header.

See vendor accuracy leaderboard →
April 2026

Crowd-sourced Early Warning Signals on every per-service page

Surfaced the existing crowd-reporting pipeline on the customer-facing status pages. Every /status/{slug} and /is/{slug}-down page now shows a "Having issues with this service?" report bar with four category buttons (Service Down, Slow Performance, Errors, Intermittent Issues) directly under the Yes/No verdict hero. Each report POSTs unauthenticated to the early-signals API and feeds: a live 24-hour spike chart with red baseline detection (Downdetector-style), a category breakdown of report types, a Leaflet density-heatmap of where reports are coming from geographically, and the threshold-based anomaly detector that fires alerts when report volume spikes above baseline, typically 10-30 minutes ahead of the vendor's official acknowledgement. Rate limited to one report per service per IP per 15 minutes. IPs are hashed with a daily-rotating salt, never stored in plaintext.

See it on a service page →
April 2026

Custom branded status page subdomains, *.status.servicealert.ai

Customers can now serve their branded status page at a free yourbrand.status.servicealert.ai subdomain, no DNS setup, no custom certificate to manage, instant activation. The wildcard TLS cert was issued via DNS-01 / Cloudflare and the nginx wildcard server block routes every *.status.servicealert.ai request through a host-based status page handler. Add or edit your status page in the dashboard, fill in the Custom Subdomain field with your chosen slug (lowercase, hyphens, 2-63 chars), and the page goes live within seconds at https://yourchoice.status.servicealert.ai. Subscribers can sign up for incident notifications directly on the page (email + RSS + Atom feeds + per-component opt-in). For full white-labelling on your own domain (e.g. status.yourbrand.com) the Enterprise plan supports custom-domain TLS provisioning.

Set up a status page →
April 2026

IndexNow integration, Bing/Yandex/Seznam re-crawl in minutes

Added IndexNow protocol support so we can push URL changes to participating search engines (Bing, Yandex, Seznam) for re-crawl within ~10 minutes instead of waiting for organic discovery. New api/lib/IndexNow.php helper exposes submit() (synchronous) and pingAsync() (fire-and-forget via register_shutdown_function) for hooking into content workflows. Priority CLI script at scripts/indexnow-submit-priority.php pushes ~128 curated URLs (homepage, tools, top 47 service pages, recent blog posts) on demand. Verification key hosted at site root. Returns HTTP 202 from IndexNow API with re-crawl confirmation.

View changelog →
April 2026

Incident Management Lifecycle with Slack war rooms and AI post-mortems

Declare incidents manually with severity levels (SEV1-SEV4) and track them through the full lifecycle: Investigating → Identified → Monitoring → Resolved. Each update posts to a timeline with optional public visibility for status page publication. Assign a Commander and multiple Responders by role (responder, communicator, observer). A new metrics dashboard shows MTTR, time-to-detect, and a 30-day severity-coded incident trend. Slack war rooms auto-create a dedicated channel per incident via the new Slack App OAuth flow, invite responders by email, post timeline updates, and archive the channel on resolve. Incident templates let you save common incident types for one-click declaration. Resolved incidents get an AI-generated post-mortem draft (summary, root cause, impact, action items, lessons learned) with a Draft → In Review → Published workflow. Public incident pages at /incident/declared/{id} show the timeline to customers linked from status pages. The severity of active SEV1/SEV2 incidents automatically updates the overall status banner on public status pages. Declare Incident button added to Slack monitor alerts to promote a monitor outage to a full declared incident in one click.

Open Incident Management →
April 2026

Granular access control, audit logs, and scoped API keys

Decide exactly what each team member can see and do, across product categories: Service Alerts, Monitors, Certificate Inventory, Certificate Posture & Compliance, Status Pages, Integrations, Team Management, Audit Logs, and Billing. Built-in roles (Owner, Editor, Certificate Manager, Auditor) cover the common cases. Business Administrators can also define unlimited custom roles like "Tier 1 SOC" or "Compliance Reviewer" and assign them to teammates with one click. Click any user row in Team Management to open a slide-out drawer with per-category None / View / Manage toggles, picked from a preset or fine-tuned per user. Bulk role operations let you select dozens of users and apply a role in one action. Multi-team membership lets a single user belong to several teams at once with per-team Lead status. Every mutation is captured in a new audit log at /audit-logs with actor, IP, timestamp, and a per-action filter for SOC 2 evidence collection, 365 days of retention for all paid plans, lifetime retention on Enterprise, exportable to CSV. API keys can now be scoped to a specific role too.

Open Team Management →
April 2026

Slack Interactive Alert Actions

Acknowledge, snooze for 1 hour, or snooze for 4 hours directly from inside the Slack message, no context switch to the dashboard. Click a button and the message updates in place to show who acted and when. Snoozed incidents stop firing duplicate alerts until the snooze window expires. Reopen restores the original buttons. Built on Slack's standard signature-verified interactivity, no extra Slack app permissions required beyond the existing webhook.

View Slack Settings →
April 2026

PagerDuty + Opsgenie Integrations

Push ServiceAlert.ai alerts into your existing PagerDuty incident pipeline or Opsgenie on-call routing. Both integrations honor your existing escalation policies, schedules, and notification rules, ServiceAlert.ai just becomes another alert source. PagerDuty uses Events API v2 with stable dedup keys so re-fires update the same incident. Opsgenie supports US and EU data regions. Recoveries automatically resolve/close the matching incident. Available on Business and Enterprise plans.

Configure Integrations →
April 2026

HSTS Preload + CAA Validation + TLS 1.3 Enforcement

Three new compliance checks on every certificate. HSTS Preload checks Google's preload list directly to determine browser inclusion status, then validates whether the served HSTS header meets preload requirements (max-age ≥ 1 year, includeSubDomains, preload directive). CAA Policy maps the cert's issuer to its canonical CAA identifier across 40+ CAs and flags violations when a cert was issued by a CA not authorized in the published CAA records. TLS 1.3 Enforcement aggregates protocol support and reports NIST SP 800-52 Rev 2 / PCI-DSS 4.0 compliance with plain-English recommendations.

View Inventory →
April 2026

Certificate Lifecycle Management, Complete Overhaul

A massive expansion of our certificate monitoring into a full Certificate Lifecycle Management (CLM) platform. New: Certificate Inventory Dashboard with portfolio analytics, Renewal Tracking with CA-specific prediction (Let's Encrypt 90d, DigiCert 398d, Amazon ACM 60d lead time), Auto-Discovery via CT logs (CertSpotter + crt.sh), Certificate Ownership with primary/secondary/team lead escalation chain, OCSP/CRL Revocation Monitoring, Policy Engine with 16 rule types and 5 prebuilt templates (PCI-DSS Baseline, NIST 800-52, Zero Trust, Let's Encrypt Best Practices, SOC 2 Type II), Multi-Environment Tracking with per-IP endpoint scanning and sync detection, Compliance Reports for PCI-DSS 4.0, SOC 2 CC7.1, and NIST SP 800-52 Rev 2 with audit-ready PDF export, Compliance Trend chart, Vendor CA Risk Monitoring (Entrust distrust alerts), Certificate Diff View showing before/after on changes, Custom Metadata Tags, and a Cryptographic Posture Score (0-100 weighted across 7 factors).

View Certificate Inventory →
April 2026

Monitor Dashboard SolarWinds-Style Redesign

Reimagined monitor dashboard with hero category bar (color-coded type counts), critical issues banner showing all down monitors with direct links, auto-refresh indicator with manual refresh button, collapsible charts section, clickable hero cards to filter by type or status, color-coded type badges throughout the list view, dense sortable table with inline 7-day uptime sparklines and uptime % column, and SolarWinds-style information density. Default view changed from grid to list.

View Monitor Dashboard →
April 2026

Type-Specific Email Notifications

Alert emails now use dedicated templates per alert type instead of forcing everything through the uptime "up → down" format. Certificate alerts use a lock icon with "Certificate Alert" subject. CT transparency alerts use a magnifying glass icon. SSL/domain expiry alerts now correctly use the certificate template. Also fixed a critical heartbeat recovery bug where parameters were in the wrong order causing alerts to crash.

Manage Alerts →
April 2026

Compliance Reports, PCI-DSS, SOC 2, NIST

Generate audit-ready compliance reports for PCI-DSS 4.0 Section 4 (6 controls), SOC 2 CC7.1 (6 controls), and NIST SP 800-52 Rev 2 (9 controls). Each control evaluated across all your certificates with pass/fail/warning status, evidence, and prioritized recommendations. Download as professional PDFs via headless Chrome. Historical snapshots tracked over time for trend visualization. Business+ plans, rate limited to 3 reports per hour.

View Compliance Reports →
April 2026

Cryptographic Posture Score

New 0-100 portfolio-level score on the Certificate Inventory page. Composite of 7 weighted factors: Grade Health (30%), Expiry Health (20%), Protocol Strength (15%), Chain Integrity (10%), Policy Compliance (10%), Ownership Coverage (10%), Revocation Status (5%). Letter grade A+ through F. Color-coded ring with progress bars showing each factor.

View Inventory →
April 2026

Monitor Dashboard Improvements

Uptime monitor dashboard now shows degraded status separately, trend arrows on uptime and alert stats, clickable stat cards, sort by status/name/response time, infrastructure tag filtering, alert channel icons and stability indicators on cards, interactive heatmap with clickable monitor names, improved sidebar with linked incidents, and persistent filter state. P95 response time line is now orange for better visibility.

View Monitor Dashboard →
April 2026

Tools Hub & Site Improvements

New /tools landing page showcasing the free DevOps and SRE tools in one place. Pricing page adds Status Pages comparison table. "Is It Down?" page enriched with FAQ schema and featured services section. Green "What's New" indicator on the Changelog nav link.

Browse Tools →
March 2026

Accessibility & Design Polish

Added semantic <main> landmarks to all pages, breadcrumb navigation on detail pages, aria-hidden on decorative SVGs, screen-reader text on status indicators, and keyboard-accessible modal close buttons. New spacing tokens, .btn-xs size, semantic badge system, and a "Recommended" label on the Business pricing card.

View Features →
March 2026

Certificate Transparency (CT) Log Monitoring

The Certificate Monitor now watches public Certificate Transparency logs for new certificates issued for your domains. Automatically detects rogue or unauthorized certificates from unexpected CAs (high-severity alert), new subdomain certificates, and routine renewals. Powered by CertSpotter with crt.sh fallback. CT entries are visible on each certificate's detail page with issuer, domain, and validity information.

View Certificate Inventory →
March 2026

Team-Shared Monitors & Group Terminology Update

Certificate Monitors assigned to a team are now visible to all team members, not just the person who created them. Added a Team column to URL Monitors and SSL Certificate lists showing which team each monitor belongs to. Renamed "Groups" to "Teams" throughout the UI for clarity.

Open Dashboard →
March 2026

Live Activity Feed Fix & WHOIS RDAP Support

Fixed the Live Activity feed incorrectly showing long-running incidents as recent status changes on every page load. Now uses server-tracked timestamps and a 24-hour cutoff for genuinely live activity. WHOIS Lookup tool now supports .dev, .app, and 18 other Google TLDs via RDAP after Google retired their WHOIS server.

Try WHOIS Lookup →
March 2026

Certificate Monitor, Dashboard Integration & Delete Support

The Certificate Monitor is now fully integrated into the dashboard. Track SSL certificate expiry, chain health, and security grades across all your domains. Added the ability to delete certificate monitors, improved the sign-in experience with benefit-driven copy, and enhanced form hints with realistic examples.

View Certificate Inventory →
March 2026

Navigation Redesign

Reorganized site navigation with a new Resources section, grouped Services dropdown with two-column layout, and streamlined Tools menu. Blog and API Docs now live under Resources alongside the new FAQ and Changelog pages.

March 2026

Dashboard Overhaul

Redesigned the dashboard with clickable service cards, improved Monitor Health cards, a new Alerts & Notifications section, enhanced empty states with guided setup, and a live activity feed with timestamps.

Open Dashboard →
March 2026

Uptime Monitoring Bundled with Business & Enterprise

Uptime monitoring is now included with Business and Enterprise plans at no additional cost. Create HTTP, keyword, ping, and port monitors to track your own endpoints alongside official status page data.

View plans →
March 2026

Hourly SLA Calculations

Uptime SLA metrics now calculate from hourly data instead of daily snapshots, providing more accurate reporting. View per-service SLA percentages on the SLA dashboard.

View SLA dashboard →
February 2026

Incident Escalation & Status.io Parsing

Improved incident detection now escalates services from "operational" to the correct severity when active incidents are found. Added Status.io parser support for services like Druva.

View incidents →