Monitor Experiments

A Flask-based web interface for comprehensive monitoring and management for the experimental setups.
Features
- Real-time status tracking for multiple experimental setups
- Advanced filtering and bulk status update capabilities
- Dynamic setup management with controlled status transitions
- Multiple authentication methods (Local and LDAP)
- User management with admin capabilities
- Active setup monitoring with visual indicators
- Remote reboot functionality via SSH
- Real-time activity monitoring with interactive plots
Status Workflow
The application enforces a strict status transition workflow:
- Running flow: ready -> running or sleeping -> running (Normal from sleeping goes to running based on the start_time of the experiment)
- Stop flow: running -> stop -> ready
- Exit flow: any_status (raised error) -> exit
Data Fields
The Control table contains the following important fields:
setup(primary key): Hostname of the machine running the experimentstatus: Current setup status (ready/running/stop/exit)last_ping: Timestamp of the last status updatequeue_size: Number of pending operationstrials: Current trial index in the sessiontotal_liquid: Total amount of reward deliveredstate: Current experiment execution statetask_idx: Index of the task to be executedanimal_id: Identifier for the test subjectip: IP address of the machine running the experiment (used for remote reboot)
Automatic Updates
The Control table is automatically updated every 5 seconds by default.
Documentation
- Setup Guide - Installation and configuration
- Environment Variables - Configuration and deployment
- User Management - Managing users and authentication
- Monitoring and Control - Monitoring setups and remote management
- Activity Monitor - Real-time activity visualization