---
title: "Monitoring and operations"
description: "Read project metrics and distinguish operation results from service readiness."
---

> Documentation Index
> Fetch the complete documentation index at: https://docs.pgnext.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitoring and operations

Open a project and **Monitoring** to read compute metrics within your authorized project scope. Select the right compute and time range before interpreting a chart.

## Read metrics in context

PostgreSQL connection counts are measurements at a point in time. Missing samples or a loading error do not mean no resource use. Do not interpret an unavailable value as zero during an investigation.

In current source, a monitoring-session CSV contains project/compute identity and sample receipt time. It reflects samples collected during the session, not necessarily complete service history or an invoice.

## Track operations

**Operations** records tasks such as resource creation or restore. Distinguish in-progress, finished and failed tasks. If a request times out, inspect both the resource and operation state before submitting again.

A finished operation does not establish that your application connected to the right SQL data. See [troubleshooting](/reference/troubleshooting/).

## Prepare a useful issue report

Record the time and timezone, action, available error code/request ID, and resource state with sensitive details removed. Do not include connection strings, passwords, MCP keys or queries containing personal data in support screenshots.

Source: https://docs.pgnext.dev/guides/monitoring/index.mdx