App Platform · Log dashboard (Grafana)

Log dashboard (Grafana)

Group several projects together and chart their logs side by side on a shared Grafana dashboard — volume by level, errors over time, and the log lines themselves.

Every project streams its own output on its Logs page. When you need to watch several of them at the same time — an API, a worker, and a frontend that all go quiet together — put them in a project group, and NevTan charts their logs side by side on a shared Grafana dashboard.

Words this page uses

Grafana — an open-source tool that turns data collected over time into charts and dashboards. Loki — the log store behind it: NevTan collects the output of your running containers into Loki, and Grafana reads it back out. Project group — a named set of your own projects, created only so their logs can be viewed together; it changes nothing about how they build, deploy, scale, or bill. Container — one running copy of a project. A project that has been redeployed a few times has had several containers, and the dashboard charts all of them, so history survives a redeploy.

Note

Nothing to install, host, or pay for — this Grafana is part of the platform and the dashboard is already built for you. The Grafana in the Marketplace is a different thing entirely: your own private Grafana instance, for your own infrastructure.

Where to find it

In the sidebar, open Developer → Project Groups. The page lists every group you have created — each card shows the group name and how many projects are in it. Click a card to open that group's combined dashboard, or New Group to make one.

Project Groups page listing group cards, each with a name and project count — Every group you have created. Click one to open its combined dashboard.

Create a group

1
Click New Group
From the Project Groups page. It opens a single form — there is no wizard and nothing gets provisioned, so creating a group is instant and costs nothing.
2
Name it
Something you will recognise later — Backend Services, Checkout flow, Staging. Up to 60 characters, and you can rename it at any time.
3
Pick the projects
Tick every project whose logs belong on one dashboard. The list holds all of your projects with a search box above it, so a long list is not a problem. At least one project is required.
Name the group, tick the projects, create — that is the whole form.
4
Create the group
You land straight on the group's page, with the combined dashboard already loading.

The combined log activity card

The group page opens on Combined log activity — a live Grafana chart of every member project's log volume, split by level, over the last hour. It is the real dashboard embedded in the console, not a picture of one, and it refreshes itself every 30 seconds.

Showing — one checkbox per project, so you can pull a noisy service out of the chart without leaving the page or removing it from the group. It appears once the group has more than one project with logs, and every project starts checked.
Open full dashboard — opens the complete Grafana dashboard in a new tab (see below). It always charts the whole group, whatever the checkboxes say.
This group's dashboard — the address the dashboard is served from, shown so you know where the tab you are about to open is going.
Projects (n) — the member list underneath, where you add and remove projects.

Note

The chart is built from what your containers write to standard output, which is what gets collected into Loki. If your app writes its logs to a file inside the container instead — a common default for Python apps — those lines still appear on the project's own Logs page, which reads them straight from the container, but they will not show up here.

Open the full dashboard

The embedded card shows one panel. Open full dashboard opens all of them in a new tab — already signed in, with no Grafana login to get past, and the dashboard's own filter row across the top.

Full Grafana Application Logs dashboard showing line counts, log volume by level, errors over time, and the log stream — The full dashboard: counts on top, charts in the middle, the log lines themselves at the bottom.

Panel	What it tells you
Containers in this view	A line at the very top mapping each project name to the raw container ids being charted — Grafana itself only ever knows the container id, so this is where the names live.
Lines / Errors / Warnings	Three counts for the selected time range. The fastest answer to "is anything wrong right now?"
Log volume by level	Lines per level over time — the panel embedded on the group page. A spike here is a change in behaviour, not just a change in traffic.
Lines by level	The same range as a bar gauge, for when you want the proportion rather than the timing.
Errors over time	ERROR, FATAL, and CRITICAL only, so a small error rate is not flattened by normal INFO traffic.
Logs	The lines themselves, newest first, with timestamps. Click a line to expand its labels and see which container it came from.

Narrowing the view

The dashboard is driven by four controls, and every panel obeys all of them at once:

Control	Effect
Project / container	Which containers are charted. The console fills this in from your group, which is why the panels are already scoped to your projects the moment the tab opens.
Level	One of `ERROR`, `WARN`, `INFO`, `DEBUG`, `TRACE`, `FATAL`, or `.*` for every level.
Search	A case-insensitive match against the line text — an order id, a stack-trace class, a path.
Time range	Grafana's own range picker. The dashboard opens on the last hour and auto-refreshes every 30 seconds.

Tip

Grafana's own container dropdown can only show raw ids like cylon-ct-6a9043c5… — it has no idea what your projects are called. Two places translate: hover a project in the Showing row on the group page for a tooltip listing its container ids, and read the Containers in this view line at the top of the full dashboard.

Managing a group

Add project — in the Projects card header. It lists every project of yours that is not already a member; adding one puts its logs on the chart immediately.
Remove — on any member row. This removes the project from the group only; the project, its deployment, and its own logs are untouched.
Rename — top right of the group page.
Delete — deletes the group and its combined view. Your projects keep running; nothing is deployed, stopped, or billed differently because of a group.

Projects card on a group page with the Add project list open, showing projects that are not yet members — Add and remove members at any time — the dashboard follows.

Access and privacy

You never sign in to Grafana yourself. When you open a group, the console asks for a short-lived signed link, and the gateway in front of Grafana exchanges it for a session as you — so the dashboard opens already signed in, and stays closed to everyone else.

A group can only ever contain your own projects, and only you can open its dashboard.
The link is valid for five minutes, and renews itself while you have the dashboard open — so a long debugging session never expires under you, while a link left lying around goes stale quickly.
The URL you land on carries no credential, so it is safe to paste into a ticket or a chat. It does not let anyone else in — they open the dashboard from their own console and see their own resources.

Troubleshooting

What you see	What it means
"Add a project below to start charting its logs here."	The group has no members yet. Use Add project.
"None of this group's projects have run a container yet, so there's nothing to chart."	The member projects have never started a container, so no logs have been collected. Deploy one and the chart fills in — see Deploy from git.
"Check at least one project above to chart its logs."	Every box in the Showing row is unchecked. Tick at least one.
"The log dashboard is not configured for this environment."	The log dashboard is switched off or not set up on this deployment of NevTan. Nothing about your project is wrong, and log collection carries on regardless.
The dashboard tab says the link has expired	The five-minute link aged out before it was opened. Go back to the group page and click Open full dashboard again — a fresh link is minted every time.
The chart is empty but the project is running	Either the time range is quieter than you expected (widen it), or the app writes its logs to a file instead of standard output — check the project's own Logs page to tell the two apart.

Logs & rollbacks

A single project's own live log stream, with level filters, search, download, and rollback.

Container metrics

CPU, memory, and network for a deployed project — the numbers alongside these log lines.

Marketplace monitoring apps

Your own Grafana, Prometheus, Loki, or Uptime Kuma instance for infrastructure outside NevTan.