Trovix Trace©
Automated root cause analysis for your infrastructure.
When something breaks — a site goes down, CPU spikes, a PHP error floods the log — Trace investigates autonomously. It collects the evidence, reasons through it with AI, and delivers a structured root cause analysis with immediate remediation steps. No manual log-hunting. No guesswork.
The problem Trace solves
When your site goes down at 2am, the investigation is manual, sequential and slow. You check the error log. You check CPU. You check the database. You check Apache. You cross-reference timestamps. By the time you know what happened, you have spent forty minutes and the incident has already resolved itself — leaving you with no documented root cause and no prevention plan.
Trace runs that investigation automatically, in parallel, the moment a trigger fires. It collects evidence from every relevant source simultaneously, sends it to an AI model that reasons across the whole picture, and produces a structured report — severity rating, root cause, immediate commands to run, and a prevention recommendation — in under 60 seconds.
How Trovix Trace© works
Detect
Trace monitors HTTP endpoints, server metrics (CPU, memory, disk), Apache error logs, PHP fatal errors, MySQL connectivity and service health. Any threshold breach or error pattern triggers an investigation.
Collect
In parallel: tail error logs, snapshot top processes, check all HTTP endpoints, query service status, read recent syslog entries. Everything timestamped and correlated.
Reason
The collected evidence is structured into a prompt and sent to an AI model. The model reasons across metrics, logs, service states and timing to identify the most probable root cause — not just the loudest symptom.
Report
A structured RCA report is written to your dashboard: severity (CRITICAL / HIGH / MEDIUM / LOW / HEALTHY), summary, root cause with evidence references, immediate remediation commands, and a prevention recommendation.
Learn
Every investigation is stored. Over time, Trace builds a record of every incident your infrastructure has experienced — what happened, why, and what was done about it.
What Trace monitors
HTTP Endpoints
Checks your URLs every 60 seconds. Detects 4xx/5xx errors, timeouts and slow responses (>5s). Fires an investigation the moment an endpoint returns an error.
Server Metrics
CPU percentage, memory usage, disk usage. Configurable thresholds. CPU is checked with a streak requirement (2 consecutive readings above threshold) to avoid noise.
Apache Error Log
Tails the Apache error log in real time. Detects PHP Fatal errors, PHP Parse errors, segmentation faults, permission denials and disk quota errors.
PHP Errors
Separate tracking of PHP fatal and parse errors from the Joomla/PHP layer. Timestamp-correlated with HTTP failures to identify cause vs symptom.
MySQL / MariaDB
Connectivity check on every investigation. Log tail for query errors, replication failures and connection exhaustion.
Service Health
Checks `systemctl is-active` for Apache, MySQL/MariaDB and PHP-FPM. A stopped service is always an immediate CRITICAL trigger.
Kubernetes / Docker
Optional. If configured, checks pod health, crash loops and recent events across namespaces.
Cloud Providers
Optional. AWS CloudWatch, GCP Monitoring, Azure Monitor can be connected via environment variables for cross-layer correlation.
The Trace dashboard
Every investigation is stored and accessible from a secure web dashboard on your own domain. No third-party SaaS. No data leaving your server. The dashboard auto-refreshes every 60 seconds and shows the full AI analysis for each incident.
Severity timeline
Every investigation rated CRITICAL / HIGH / MEDIUM / LOW / HEALTHY. At a glance, see whether your infrastructure health is trending better or worse.
Full AI analysis
Click any incident to read the complete structured root cause analysis — what broke, why, what to do now, and how to prevent recurrence.
Live metrics
Latest CPU, memory and disk readings from the most recent investigation displayed at the top of the dashboard.
Manual investigations
Trigger an investigation on demand from the command line whenever you suspect a problem — even if automated thresholds haven't fired yet.
Trace in the Trovix product family
Trovix Audit©
Where Trace investigates infrastructure incidents, Audit monitors the performance and governance of your AI systems. Together they cover the full operational stack.
Open Audit© →Trovix Watch©
Watch monitors regulators and policy changes at the firm level. Trace monitors your infrastructure at the server level. Both alert you before a problem becomes a crisis.
Open Watch© →OpsPilot & AgentOps
Trace is built on the same agent architecture as the OpsPilot and AgentOps modules in the Trovix AI catalogue. Those modules extend the concept to enterprise Kubernetes and multi-cloud environments.
Browse AI catalogue →Stop hunting through logs at 2am.
Trace installs in 20 minutes and investigates your first real incident before the end of today. Setup call not required — but available if you want one.
