Built-in anomaly detection and alerting.
Complex environments require adaptive tools. Flowtune triggers alerts when behavior deviates from the per-service baseline profile, avoiding false alarms from misconfigured fixed thresholds. Webhook integration enables automatic mitigation.
Failure can be localized to one region, cluster/availability zone or IP/instance. Understand what part of your service and your upstream and downstream providers is affected for the most effective mitigation. Gather the required evidence to initiate inter-region failover with confidence.
One second resolution.
Get faster, lower-noise alerts by using high-resolution, one-second metrics. Even short blips of 10-30 seconds are extremely visible to users, and are surprisingly common with some large providers.
No app code changes, 100% visibility.
(see "HOW IT WORKS" below)
HOW IT WORKS
Flowtune employs an agent: a self-contained binary that is executed on each instance/VM/bare-metal machine. The agent collects a detailed “flight recording” of all service communication directly from the operating system with very low overhead (< 0.1% CPU and network), requiring no modification to applications. The agent performs partial aggregation and sends one-second digests to the processing backend, which performs further aggregation, anomaly detection, alerting, and visualization.
SUPPORTED OPERATING SYSTEMS
RHEL, CentOS, Amazon Linux, Debian, Ubuntu
Linux kernel versions 4.4 - 4.14
Flowtune was founded in 2016 and is based in Sunnyvale, CA. The technology is based on 6+ years of research at MIT. The Flowtune Service Monitor is deployed in production in a large unified communications company ($2B+ market cap).
250+ INSTANCES IN YOUR DEPLOYMENT?
Contact us to try it out!
Flowtune, Inc. | 100 S Murphy Ave #200 | Sunnyvale, CA 94086