25.03 Media Server
In the March release, we expanded monitoring functionality: added a dashboard with the bitrate of incoming streams, the ability to share statistics with colleagues, and set up alerts to avoid missing incidents. If you want to get complete information about the quality of incoming signals and identify the causes of interference for viewers, contact our managers. They will explain how to connect the monitoring service.
Let’s take a closer look at the other updates.
Configure alerts in monitoring to avoid missing critical incidents
Now you can set up custom alerts when a counter reaches a certain value. Notifications are implemented using Grafana’s standard tools.
To add an alert, click on the “plus” icon in the upper right corner and select New alert rule.
Notifications can be sent via email or another service of your choice, such as a Telegram bot, which you can add during setup.
Share access to statistics with colleagues
In the new release, users added to your Media Server account as colleagues will be able to view statistics from the Retroview tab, including the new monitoring feature. Now, it’s easy to share data about viewers and stream quality with managers, executives, and technical specialists.
Track the root causes of problems down to a specific PID
We have divided error counters into channel-level errors and errors within a specific Program ID (PID). Now, you can analyze error statistics in monitoring tool down to individual tracks.
Use new monitoring statistics: source bitrate dashboard and lost PID counter
We’ve added a dashboard displaying the bitrate of all incoming channels (including NDI) and its changes over time. A new counter, errors_pid_lost, has also been introduced. It shows the number of lost PID from MPEG-TS or lost tracks from other sources.
Play LL-HLS archive without switching to a new playlist
Previously, when viewers rewound LL-HLS streams, Media Server switched to HLS protocol for archive playback, which required time to download the HLS playlist.
In the new version of Media Server, we have added the ability to play the current LL-HLS stream archive without switching to HLS, using event playlists (/rewind-3600.ll.m3u8
, /archive-xxxxxxxxxx-now.ll.m3u8
). This significantly speeds up archive playback for viewers.
Reduced the number of Internal Server Errors when reading the archive
Simultaneous execution of multiple recording processes could cause segment corruption in dvr_cache, leading to 500 errors. To reduce server load, we implemented serialized recording in a single process, which significantly decreased the number of failures when playing the archive.
Monitor archive access statistics
We have added statistics that help track how often viewers access newer or older archive segments. This data is currently available via the API in the stream_get operation.
Use the method dvr_read_popularity to obtain archive reading statistics for today, yesterday, the past week, and earlier periods.
In upcoming updates, we will introduce monitoring of archive access statistics. This will help better understand user behavior and efficiently calculate storage capacity for edge servers.