Use case · Software, automation, operations
Anonymised, like all our use cases
Broadcast operations, monitored around the clock.
For every broadcast day, a TV channel needs a playlist, programme data and an as-run log, spread across several servers. For a group of TV channels we rebuilt the monitoring: it checks all files and live streams continuously, reports every new problem exactly once and tells a missing file apart from a server that is down.
Deployed in
TV channel group
Services
Software, automation, operations
Scope
6 channels, 8 live streams
Data sources
6 servers via FTP and SFTP
The software checks. The team acts.
Every file has a deadline. Before it the file is pending, after it overdue, and only then does exactly one alert email go to the people responsible. Once the file arrives, the issue is closed.
How a check works.
- Fetch A background service queries the channels’ servers every 10 minutes and the upload folders every 5 minutes.
- Check Rules per file type: playlists for the next three broadcast days, age of the programme data, as-run logs for yesterday and the day before.
- Assess Before the deadline a file is pending, after it overdue. If a server does not respond, the status is “source unavailable”.
- Streams Every live stream is checked once a minute: online, frozen or unreachable.
- Alert A new problem triggers exactly one email to the responsible recipients. Once a day a report lists everything that is still open.
10 min
Interval at which the channels’ servers are queried
60 s
Interval at which every live stream is checked
1 email
per new problem, no repeats
90
days of history for all checks, outages and emails
The status picture for the wall.
On a second screen the team sees at a glance whether everything is on track: open issues, the next deadline, the channels live and on-time delivery over the last 14 days. View only, refreshed every 15 seconds.
One application, six areas.
From the file on the server to the picture on air in one place, with roles for the team and administration.
Software
Status board
One row per channel: playlists, programme data and as-run logs. “Check now” fetches a fresh status immediately.
Automation
Ingest
Uploads, import into the playout system and finished files in lists. Failed imports and missing files come first.
Software
Channels live
All live streams in one view, with their status, since when, and the outages of the last seven days.
Software
Wall dashboard
Full screen for a second display, light or dark, refreshed every 15 seconds. With on-time delivery for the last 14 days.
Automation
Alert emails
Recipients by file name, personal alerts per event and a daily report of everything that is open.
Platform
Administration and roles
The team maintains channels, streams, deadlines and recipients itself. One login for every application on the platform.
What makes the difference.
No false alarms
If a server does not respond, the software reports neither “missing” nor “all fine”. If it stays down for 30 minutes, a separate alert goes out.
Deadlines, not constant red
A file that is only due tomorrow is pending, not an error. The team maintains the deadlines.
One email per problem
No inbox full of repeats. Whatever stays open is collected in the daily report.
Short dropouts stay quiet
A stream alert only goes out after three checks in a row without a picture, so after about three minutes.
What we built it with
- .NET 10
- Blazor
- .NET Aspire
- PostgreSQL
- Redis
- FluentFTP
- SSH.NET
- MailKit
- Docker
- GitLab CI
- Ansible
- OpenTelemetry
What must never fail unnoticed at your company?
In an intro call we look at which of your processes can be monitored and automated.