Section 04 — The pre-emptive Q&A
The six questions your staff engineer will email us about anyway.
Asked by SRE leads at Rippling, Mercury, Linear, and a few hundred Series B+ engineering teams we can't name yet. Short answers, no marketing fog.
- Is CoreWatch a drop-in replacement for the OpenTelemetry Collector, or do I have to re-instrument?
- Drop-in. Point your existing OTLP exporters at our endpoint — the collectors, SDKs, and auto-instrumentation agents stay unchanged. We've been OTLP-native since day one (the founders wrote the first OTel Collector releases at Google). No proprietary agents to install, no sidecar gymnastics, no re-instrumentation.
- What does the ingest path actually look like under load?
- OTLP gRPC → load balancer → ingest gateways → Columnar Adaptive Index on local NVMe → S3-compatible cold tier. Sustained 11.4M spans/sec/node on commodity c6id.4xlarge. End-to-end p99 query latency on a 30-day window is under 1.4s for a 100M-span dataset.
- How does retention and billing actually work — am I going to get a surprise invoice?
- You're billed on hosts × hours and spans × volume — both visible on a live usage page down to the minute. No custom-metrics tax, no high-cardinality tag surcharge, no per-user seat fee. If you cross your tier's span cap, ingestion throttles with a banner; it doesn't auto-upgrade you. Caps and overage math are documented publicly.
- What's the migration path off Datadog / New Relic without a 3-month dual-pay window?
- Run CoreWatch in parallel for the first 30 days at no charge (Team tier trial, billed only after day 31). Cut over one service at a time using OTLP routing — same SDK, different endpoint. We provide a side-by-side query translator for Datadog NRQL and New Relic NRQL so your runbooks port over verbatim. Most teams complete migration in 4–6 weeks.
- Is the HIPAA tier real, and what's actually covered?
- Yes — HIPAA-compliant logging has been generally available since March 2023, on Enterprise only, with a signed BAA. Covers PHI redaction at ingest, encrypted PHI fields with customer-managed KMS keys, and a separate audit log channel. Not FedRAMP (we don't claim it — Datadog GovCloud and Splunk FedRAMP exist for that).
- Do you have a status page, and what's your uptime been on the ingest path?
- Yes — status.corewatch.net. 99.97% measured ingest-path uptime over the last 12 months. Team SLA is 99.9%, Enterprise is 99.99%. Incidents are written up as post-mortems on the same status page within 48 hours, signed off by an SRE director.