Recherche de notes de version
documents correspondants
Aucun document correspondant n'a été trouvé
23.11 Central
In Flussonic Central 23.11, we are pleased to announce a series of noteworthy enhancements that add depth and sophistication to our platform. Transitioning away from previous methods, we now exclusively employ Central cluster ingest technology when incorporating streamers into the Central-managed cluster. Additionally, we extended flexibility to the creation of archive fragment URLs, refined video analytics node requests, and instituted a safeguard mechanism to prevent stream overuse. The introduction of tags for streams and streamers grants users enhanced control over their allocation strategies. And that’s not all, read the 23.11 release notes to learn about the freshest updates we made to Central that improves your user experience!
Cluster ingest
To allow users to leverage all the benefits and advantages offered by Central cluster ingest technology, the previous method of configuring streamers using peer configurations for cluster ingest has been discontinued. When adding a streamer to a cluster that is managed by Central, the streamer now utilizes Central cluster ingest technology exclusively.
URLs of archived fragments
To adapt the system for a wider range of scenarios we gave our customers a possibility to create archive fragment URLs with a greater flexibility. The Public Payload URL or API URL can now be placed before the hostname in the URL structure, which is especially helpful when dealing with hostnames that are not FQDN-compliant.
Video analytics nodes
To improve resource utilization and streamline video analytics operations with Central, we optimized the number of requests sent to offline nodes. This update enhances the Central efficiency by adding a check to verify the online or offline status of video analytics nodes before attempting to fetch episodes from them.
Number of streams
For better resource allocation and preventing overuse of server capacity, we ensure the system operates within a defined number of streams. This update adds a safeguard within Flussonic Central to check and prevent the creation of streams beyond the specified channel_limit for a given streamer hostname.
Stream layout history
To allow users to configure their streams and streamers more effectively and to minimize layout changes, we added the ability to get a stream layout history showing the reason of the change in arrangement. This helps to understand why a stream moves one streamer to another, how often this happens, as well as configure streams and streamers to minimize layout changes.
Streaming_endpoint parameter
To allow the use of Watcher Client UI on VSAAS.IO and other cluster installations, the use of streaming_endpoint parameter was implemented to receive a link for the playback from the server where the video from an IP-camera is currently ingested.
To enhance the flexibility and control over stream allocation and management in Central, optimizing resource usage and stream distribution within the cluster, a new feature was introduced in Central, allowing users to manage tags for both streams and streamers. These tags provide more control over how streams are allocated when adding, replacing, upgrading, or removing streamers within a cluster. Users can use tags to determine server priorities for reallocating streams in case of a failure, or they can assign specific sets of streams to particular groups of streamers in the cluster.
Configuration parameters
To offer valuable insights into the streamer status, operational history, and unique identifiers, enhancing the ability to monitor and manage streamers effectively, we included additional parameters such as uptime, [id] (https://flussonic.com/doc/api/central/#tag/config/operation/config_get/response%7Cstats%7Cid), and started_atthat are returned in the GET /config request in the recent update to Central API.
To enhance the observability and efficiency of Flussonic Central within larger-scale server deployments, facilitating better performance analysis and alerting through integration with Prometheus, the ability to collect Central metrics via the “/config” endpoint was added in OpenMetrics format.
To help customers with monitoring the status and capacity of Flussonic Central cluster, facilitating effective management and performance monitoring of their media streaming infrastructure, we enabled the feature of retrieving streamers’ statistics as part of the response when making a specific GET request to the "/config" endpoint.
To facilitate effective monitoring of the cluster status and capacity, contributing to the efficient management of their media streaming infrastructure, we empowered Flussonic Central with the feature of retrieving agent statistics as part of the response to a GET /config request.
To retrieve and access the ’listener’ configuration settings through this API endpoint, providing more comprehensive information about the configuration of the system, the configuration related to “listeners” is now included in the response when making a GET /config request .
Streamer polling interval
To reduce the load on the Central server, the default streamer polling interval was decreased from once every two seconds to 30 seconds. Additionally, administrators now have the option to further customize the polling interval by adjusting the OBSERVER_POLLING_INTERVAL configuration parameter to suit their system performance and monitoring needs.
Streams reallocation
For an optimal use of the resources within the Flussonic Central environment resulted in delivering high-quality media streaming services, a dynamic stream reallocation feature was implemented in 23.11. When a new streamer is added to the system, the platform will assess whether reallocating streams to different components will lead to improved overall system performance.
Merging Central and analytics
To reduce the load on the Flussonic Central server the authentication process for analytics-related operations is simplified. Instead of requesting permissions from Central, the analytics nodes can now authenticate themselves using the allow_tokens parameter. This results in more efficient and streamlined authentication for analytics tasks within the Flussonic Central system.
Streamer hostnames
With 23.11 we added greater flexibility in specifying hostnames, particularly for cases where domain names with a trailing dot are used. Flussonic Central has made a change to its validation rules for streamer hostnames, it no longer enforces the rule that hostnames cannot end with a dot (.).
API key
To enable Vision servers to operate in a cluster managed by Central while providing the necessary authentication and access control through the API key, we added a feature that allows users to specify the API key of analytics nodes in the API_URL parameter in the format http://api-key@hostname.
Bugs fixed
Spontaneous restart of Central
There is no longer a problem of spontaneous restart of the system with “panic” errors in the logs. This ensures that the Flussonic Central system operates reliably and without unexpected disruptions, providing a more stable and dependable management and monitoring platform for Flussonic Media Servers.
Status of streams
Since 23.11 Flussonic Central customers can obtain accurate and up-to-date information in stats about the online/offline and provisioned/not provisioned status of streams within their cluster configuration. This improvement aids in effective monitoring and management of the streaming infrastructure.
Channel_limit parameter
The problem of the channel_limit parameter being periodically and unexpectedly reset was resolved.. This ensures that the parameter remains stable and consistent, allowing administrators to manage the maximum number of channels or streams effectively within their Flussonic Central system.
Agents
Streams associated with Agents can be effectively managed and run without encountering conflicts or issues that were previously leading to the “keeplive drop” error message. This improves the reliability and performance of stream management within the Flussonic Central system.
Episodes
When making GET requests for stream information associated with specific hostnames, Central users do not encounter the problem of receiving incorrect “episodes_expiration” data when making GET requests for stream information associated with specific hostnames. This ensures the accuracy and reliability of the information retrieved from the Flussonic Central system.
Double slash in URL
The URLs returned by Central balancer for redirects no longer contain double slashes. This correction ensures proper URL formatting and functionality, preventing potential issues that may arise from incorrectly formatted URLs, such as routing problems or broken links.
Person management API
The fix ensures that users or applications interacting with this API experience faster and more responsive cursor-based navigation and data retrieval in the person management API
23.11 update ensures that the HTTP 404 error no longer occurs when using the person management API in Flussonic Central. The issue has been fixed, and the API should now respond as expected, without returning “Not Found” errors for valid requests.
Analytics of events
This update addresses and resolves problems related to HTTP 500 errors and timeouts during event collection from Flussonic Central. By fixing these issues, it ensures that events are collected without disruptions, reducing the risk of missed analytics events and contributing to the overall reliability of the Flussonic Central system.
Central Admin UI
An issue where streams were not playing as intended in the player located on the Overview tab of the Central Admin UI was fixed. As a result of this fix, users and administrators should now be able to view and play streams within the UI without encountering playback issues.
PostgreSQL logs
The system now operates more smoothly without the PostgreSQL log-related disruptions, ensuring that events are received and processed correctly.
Location of archive parts
Outdated information about the archive parts location on streamers is removed, and unnecessary inputs are no longer provisioned, which improves the system efficiency and avoids unnecessary load on the system.
Large number of streams
For the memory resources to be managed more efficiently, contributing to improved service stability, the system defense was strengthened against memory leaks that may occur during the connection of Agents.