SiftLog

M Media Software

About this app

It's 2am. Your Infrastructure. In Your Pocket.

SiftLog runs on your server, watches your logs continuously across every service, and detects cascade failures, anomalies, and silence the moment they happen - with zero performance impact on your infrastructure. Every monitoring tool tells you something is broken. SiftLog tells you what broke first. This free app delivers those signals directly to your phone - over your own LAN or VPN, no cloud in the middle.

When your infrastructure has a problem at 2am, you need to know three things before you open your laptop: what broke, which service broke first, and whether it's still happening.

SiftLog Platform Monitor puts that on your phone.

WHAT YOU SEE

• Cascade chains - which service failed first and what it took down with it
• Anomaly alerts - error rates spiking above the rolling baseline
• Silence detection - services that stopped logging entirely, a failure mode that manual log review consistently misses
• Per-service health status across all your sources
• The full time-ordered event stream with severity color coding

NOTIFICATIONS THAT MEAN SOMETHING

Most monitoring alerts tell you something is broken. SiftLog tells you what broke first. When your phone buzzes at 2am you see:

Cascade Detected
auth-service -> api-gateway -> payment-service
Detected 2 minutes ago

That is the incident. One line. You know where to start.

RUNS ON YOUR INFRASTRUCTURE

SiftLog is self-hosted. No agents. No sidecars. No changes to your services or log format - it reads from the log infrastructure you already have. Your logs never leave your server. The app connects directly to your daemon over your local network or VPN - no cloud relay, no third-party data handling.

The android app is free for your entire staff. No subscription, no tiers, no upsell.

The REST API is off by default. You enable it, set an API key, and the app connects with that key. If you never enable it, there is never an open port and no network exposure whatsoever.

HOME SCREEN WIDGET

One health dot per service. All green, you're fine. One goes red, tap it -- before you're fully awake, you know whether it needs you. The 2x1 widget shows dots only. The 4x2 widget adds the last signal and the time it was detected.

MULTIPLE ENVIRONMENTS

Add production, staging, and any other SiftLog instances from Settings. All signals merge into a single time-ordered feed so you see the full picture across environments in one place.

REQUIREMENTS

This app requires a running the SiftLog daemon. It connects to your daemon's REST API over your local network or VPN. It does not function as a standalone application.

SiftLog daemon: zero instrumentation, fully self-hosted, supports Loki, CloudWatch, local files, syslog, journald, and TCP streams.

https://www.mmediasoftwarelab.com/product/siftlog-platform/
https://www.mmediasoftwarelab.com/siftlog/platform/

What’s new

Connect to your SiftLog daemon and see what it knows.

• Cascade chains — which service failed first, what it took down
• Anomaly alerts — error rates above the rolling baseline
• Silence detection — services that stopped logging
• Per-service health status and full event stream
• Home screen widget with one dot per service
• Multiple environments in a single feed

Connects directly to your daemon over LAN, VPN, or Tailscale. No cloud relay. No account required.