Introduction
Every monitoring screen in NoSqlStudio — mongostat, mongotop, Current Op, Replica Set Status, the Realtime dashboard — polls your server while its tab is open. Close the tab and the polling stops. The background recorder changes that: it turns a recording into a job that keeps running on its own, independent of any tab.
What this is for
You start a recording on the screen you're already looking at, then you're free to close that tab, open other tabs, or just leave NoSqlStudio working in the background. The recording carries on. When you come back, the screen's History view shows everything that was captured while you were away.
Concept
What the recorder is
A recording is a small background job. It polls one connection on a fixed cadence and writes each sample to the same history store the screen's own History view reads. It is owned by the recorder engine — not by the tab — so closing the tab does not stop it.
Heads-up
This is a workstation tool, not 24/7 observability
The recorder runs on your machine, inside NoSqlStudio. If you close the app, log off, or the machine sleeps or loses power, capture stops — there is no agent on the server. Use it for focused investigation windows (an incident, a load test, a suspicious nightly job), not as a replacement for server-side monitoring.
How to read this guide
Each step has:
- Do — the exact action.
- See — what should happen on screen. This is your pass / fail check.
- Why — background, when it helps. Skip it if you're in a hurry.
You will need a connected MongoDB connection and a monitoring tab open on it.
Start a recording
A recording always starts from a monitoring screen — the one you want to capture.
Open a monitoring screen
Open any of the recordable screens — mongostat, mongotop, Current Op, Replica Set Status, or the Realtime dashboard — on a connected connection.
The live screen, polling as usual. In its toolbar there is a 🔴 Record button.
Click 🔴 Record
Click the 🔴 Record button in the toolbar.
The button changes to ⏹ Stop recording and takes on a red outline. A background recording is now active for this connection.
The button reads the live recorder state. If the recording ever stops for another reason — you stop the job from the Job Manager, or the connection drops — the button flips back to 🔴 Record on its own.
Concept
Shortcut from the sidebar (database-scoped)
Right-click any database in the left sidebar and pick 🔴 Start recording (Profiler). Same outcome as opening the Profiler manually and clicking Record: the Profiler tab opens with that database pre-selected, and a database-scoped recording starts in the background — in a single click.
Choose the cadence (optional)
Before recording, set the screen's refresh interval to the cadence you want captured. The recorder polls at the interval shown when you press Record.
Concept
A fast interval (250 ms) is great for a live screen but heavy for a long recording. For a recording you'll review later, a slower cadence — a few seconds — keeps the history compact without losing the trend.
See it in the Job Manager
A recording is a job, so it shows up everywhere jobs do — visible and controllable.
Open the Job Manager
Open the Job Manager tab. The filter dropdown at the top right has a Monitoring (N) option that narrows the list to just recording jobs.
A Monitoring job for the connection you just started recording. It shows the connection, the monitoring type, and the scope being recorded.
The Job Manager is the single place that lists every running recording. A DBA who opened several monitoring tabs can tell at a glance exactly what is — and is not — being recorded.
Concept
Watch (Change Watcher) also counts as a Monitoring job
A live Watch with its own recording toggle on registers in the Job Manager under the same Monitoring filter — there is no separate Record button on the Watch screen because the change-watcher already owns the stream + the history mirror. The recorder just surfaces it here so the DBA sees every active recording in one place.
Check the sidebar — recording dots
Look at the connection in the left sidebar.
A connection with an active recording carries a small red dot that pulses next to its name. The dot appears at every level the recording covers: on the connection for any active recording, on a database when a database-scoped recording (e.g. Profiler) is running, on a collection when a collection-scoped one is.
Concept
What the dot colour means
Red (pulses every ~1.6 s) — recording is healthy, samples are landing in the management database. Amber — the management database is briefly unreachable, so samples are buffering to a local file (a short outage; samples will be reconciled when the database comes back). Red, fast pulse — the recording needs your attention (lost connection, re-auth needed).
Concept
Filter the sidebar to only recording connections
Click the funnel icon at the top of the sidebar to open the filter popover. A Background recording group lets you narrow the list to All connections (default), Only recording, or Only not recording. Useful when you have many connections open and want to focus on the ones with active capture — or the ones without.
Close the tab — capture continues
This is the whole point of the recorder: the recording is decoupled from the tab.
Close the monitoring tab
Close the monitoring tab you started the recording from. Open other tabs, or just keep working.
The Job Manager still shows the Monitoring job, still running. The sidebar dot is still red. Nothing was interrupted.
The recorder engine is a single app-level service. It owns the poll loop. The tab was only ever a viewer — closing it removes the viewer, not the recording.
Reopen the screen and review
Reopen the same monitoring screen and switch it to the History view.
The history covers the whole window — including the time the tab was closed. The gap you'd normally see is gone.
Where the recorded data goes
The recorder reuses each screen's existing history store, so recorded samples live exactly where the screen's own History view already looks.
Management database, primary
Configure a management database in Settings ▸ Database if you haven't already.
Recorded samples are written to that database. It is the primary store — durable, and shareable across machines.
Local disk, contingency only
Concept
If the management database is briefly unreachable, samples buffer to a local file so nothing is lost. When the database comes back, the buffered samples are reconciled into it. Disk is the fallback, not a parallel copy — when the database is healthy, the recorder writes only there.
Honest limits
Knowing what the recorder does not do is as important as knowing what it does.
The machine has to be on
Concept
Across an app restart, recordings auto-resume
What the recorder gives you back is continuity around the restart. Every Record you pressed and didn't Stop is remembered. When you reopen NoSqlStudio, each recording automatically resumes as soon as its connection reconnects — under a fresh Job Manager entry. You don't need to re-press anything. The recording from before the quit closes as Done (interrupted) with a timestamp, and a new one continues from there.
Reconnection and re-authentication
If a connection drops, the recording job reconnects on its own as long as its credentials can be renewed without a prompt (SCRAM, X.509, AWS-IAM).
Recording is per connection and scope
Pressing Record on a screen records that connection, at the scope shown on the job. Opening other monitoring tabs does not record them — each recording is explicit. What the Job Manager lists is exactly what is being captured: no more, no less.
Stopping a recording
A recording runs until you stop it. There are two equivalent ways:
- On the monitoring screen, click ⏹ Stop recording in the toolbar.
- In the Job Manager, stop the Monitoring job.
Either way the poll loop ends, the job is marked done, and the sidebar dot clears. Disconnecting the connection also stops every recording on it — a job cannot poll a server it is no longer connected to.
Concept
Recordings are deliberately explicit to start and explicit to stop. There is no hidden capture — if you don't see a Monitoring job, nothing is being recorded.
Summary
| Question | Answer |
|---|---|
| How do I start a recording? | Click 🔴 Record in any recordable monitoring screen's toolbar. |
| Does it keep running if I close the tab? | Yes. The recording is a background job, independent of the tab. |
| Where do I see active recordings? | The Job Manager (Monitoring jobs) and the sidebar recording dot. |
| Where is the data stored? | The configured management database (primary); local disk as a fallback during outages. |
| What happens if I close the app? | Capture pauses while the app is off. When you reopen, every recording auto-resumes as soon as its connection reconnects — no re-pressing needed. |
| What does the sidebar dot mean? | An active recording: red = healthy, amber = mirror briefly unreachable (buffering to disk), red fast pulse = needs attention. The dot appears at the level being recorded — connection, database, or collection. |
| How do I stop it? | ⏹ Stop recording on the screen, or stop the job in the Job Manager. |