Skip to main content
The Monitoring feature provides real-time visibility into database server activity. View active queries, kill long-running processes, and track connection counts and server health.

​
Opening the Monitor

1

Access Monitoring

Click the connection context menu → Monitor, or click the monitoring button in the sidebar.
2

View Process List

A table displays all active database processes (connections) with query details.
3

Enable Auto-Refresh

Toggle Auto-Refresh to update the process list every 3 seconds.
Monitoring is available for PostgreSQL, MySQL, MariaDB, MongoDB, Redis, and SQL Server. SQLite and DuckDB are embedded databases without active processes.

​
Process List

​
Process Table Columns

The process list displays the following information:
ColumnDescription
IDProcess or connection ID (unique identifier)
UserDatabase user running the query
DatabaseActive database for the connection
CommandQuery type (SELECT, INSERT, UPDATE, etc.)
TimeHow long the query has been running (seconds)
StateCurrent state (active, idle, waiting, etc.)
QueryThe SQL query being executed (truncated to 100 characters)
ActionsKill process button

​
Color-Coded Time

Query execution time is color-coded for quick identification:
  • White/Default: Less than 1 minute
  • Amber: 1-5 minutes (potentially slow query)
  • Red: Over 5 minutes (long-running query, potential issue)

​
Sorting Processes

Click column headers to sort the process list:
  • Sort by Time: Identify long-running queries
  • Sort by User: Group queries by user
  • Sort by Database: Filter processes by database

​
Auto-Refresh

Enable Auto-Refresh to monitor processes in real-time:
1

Toggle Auto-Refresh

Click the Auto-Refresh toggle in the status bar (or in the monitoring toolbar).
2

Live Updates

The process list reloads every 3 seconds automatically.
3

Disable

Toggle Auto-Refresh off to stop automatic updates. Manually click Refresh to reload.
Auto-Refresh pauses when you switch to a different tab to reduce server load. It resumes when you return to the monitoring tab.

​
Server Status

The status bar displays server health metrics:

​
PostgreSQL Metrics

  • Active Connections: Current number of client connections
  • Max Connections: Maximum allowed connections (from max_connections setting)
  • Connection Usage: Visual indicator (e.g., “45/100” with progress bar)

​
MySQL/MariaDB Metrics

  • Threads Connected: Current active connections
  • Max Connections: Maximum allowed connections
  • Uptime: Server uptime
  • Questions: Total queries executed since startup

​
MongoDB Metrics

  • Current Connections: Active client connections
  • Available Connections: Remaining connection slots
  • Uptime: Server uptime

​
Redis Metrics

  • Connected Clients: Active client connections
  • Max Clients: Maximum allowed clients
  • Used Memory: Current memory usage
  • Uptime: Server uptime
Server status metrics appear in the status bar at the bottom of the monitoring tab.

​
Killing Processes

​
Kill a Single Process

1

Identify Process

Locate the process you want to terminate in the process list.
2

Click Kill

Click the trash icon in the Actions column.
3

Review Details

A confirmation dialog shows process ID, user, database, and the query being executed.
4

Confirm Kill

Click Kill Process to terminate it immediately.
Killing a process terminates the connection and rolls back any uncommitted transactions. Use with caution.

​
Force Terminate (PostgreSQL)

For PostgreSQL, choose between gentle cancellation and forced termination:
Uses pg_cancel_backend() to gracefully cancel the current query. The connection remains open.Use when: You want to stop a long-running query but keep the connection alive.
Enable Force terminate in the kill confirmation dialog for PostgreSQL.

​
Common Kill Scenarios

Identify the query with the highest Time value (red highlight). Kill it to unblock waiting processes.
Look for processes with state “idle in transaction” that have been running for a long time. Kill to release locks.
If you accidentally ran SELECT * FROM huge_table without a LIMIT, kill the process to stop it.
Identify connections from misbehaving applications (check User and Database) and kill them.

​
Troubleshooting

  • Ensure you’re connected to the database
  • Check that the database user has permission to view processes:
    • PostgreSQL: SELECT pg_stat_activity requires pg_monitor role
    • MySQL: Requires PROCESS privilege
    • MongoDB: Requires inprog command access
  • Ensure the database user has permission to kill processes:
    • PostgreSQL: Can only kill your own processes unless you have pg_signal_backend role
    • MySQL: Requires SUPER privilege to kill other users’ processes
    • MongoDB: Requires killop privilege
  • Check Auto-Refresh is enabled
  • Verify the database connection is still active
  • Click Refresh manually to test
  • Queries are truncated to 100 characters in the table
  • Click on the row to view the full query in the right panel
  • For very long queries, copy from the kill confirmation dialog

​
Privacy Mode

When Privacy Mode is enabled:
  • User column is blurred
  • Database column is blurred
  • Query column is blurred
This prevents exposing sensitive information when screensharing. Enable Privacy Mode: Click the eye icon in the status bar or press Shift+Cmd+P.

​
Keyboard Shortcuts

ShortcutAction
Cmd+RRefresh process list
Shift+Cmd+PToggle Privacy Mode

​
Database-Specific Notes

​
PostgreSQL

  • Process list fetched from pg_stat_activity view
  • Can kill processes with pg_cancel_backend() or pg_terminate_backend()
  • Shows queries from all databases on the server

​
MySQL/MariaDB

  • Process list fetched from SHOW FULL PROCESSLIST
  • Killed with KILL <id> command
  • Requires PROCESS privilege to view all processes

​
MongoDB

  • Process list fetched from db.currentOp() command
  • Killed with db.killOp() command
  • Shows operations on the current database

​
Redis

  • Process list fetched from CLIENT LIST command
  • Killed with CLIENT KILL command
  • Shows all connected clients

​
SQL Server

  • Process list fetched from sys.dm_exec_requests and sys.dm_exec_sessions views
  • Killed with KILL <session_id> command
  • Requires VIEW SERVER STATE permission

​
Use Cases

Open the Monitor to identify slow queries causing application lag. Sort by Time to find queries running for minutes.
Monitor active connections over time. If connections grow without being released, your application may have a connection leak.
If a deployed migration script is running forever, use the Monitor to kill it before it locks the database.
Check Active Connections vs Max Connections to determine if you need to increase connection limits.

​
Best Practices

Keep the Monitor open during deployments to watch for long-running migrations or unexpected queries.
If active connections consistently approach max connections, consider:
  • Increasing max_connections
  • Using connection pooling (PgBouncer, ProxySQL)
  • Investigating application connection leaks
If you frequently see the same slow query, document it and optimize with indexes or query rewrites.
Enable Auto-Refresh when troubleshooting live issues. Disable it during normal development to reduce server load.

Build docs developers (and LLMs) love

Get started for freeTalk to us