What is GSLB? Global Server Load Balancing

Share on facebook
Share on twitter
Share on linkedin
Ceeblue GSLB: Global Server Load Balancing. Load Balancing, Low Latency, High Availability, Persistent Ingest

Global Server Load Balancing (GSLB) is a critical technology that enhances the reliability, performance, and resilience of networked services. By intelligently distributing traffic across multiple servers located in diverse geographical Points of Presence (PoPs), GSLB ensures that users experience reduced latency, increased uptime, and seamless service even during failures or high traffic periods.

This article explores the intricacies of GSLB, its benefits, and its specific applications in the streaming industry, focusing on how Ceeblue’s specialized GSLB solutions optimize stream ingestion and enhance user experience.

What is GSLB?

Before diving into specifics, let’s clarify the concept of a Global Server Load Balancer (GSLB). A GSLB is an intelligent system that manages the distribution of network traffic across multiple servers situated in Points of Presence (PoPs) around the world.

The benefits of GSLB include:

  • Increased reliability and reduced latency
  • Improved performance by passing user requests to the nearest servers
  • Disaster recovery and high availability to minimize the impact of data center or network failures

In a typical GSLB setup (Figure 1), a client sends a DNS query to a DNS resolver for a specific hostname. The DNS resolver then forwards this query to a service, which selects an optimal server IP address based on factors like load (health) and network conditions. This IP address is then returned to the DNS resolver, which finally responds to the client with the resolved IP address.

A typical GSLB setup: a client sends a DNS query to a DNS resolver for a specific hostname. The DNS resolver then forwards this query to a service, which selects an optimal server IP address based on factors like load (health) and network conditions. This IP address is then returned to the DNS resolver, which finally responds to the client with the resolved IP address.

Benefits of Global Server Load Balancing for Stream Ingestion

In the context of streaming, low latency and resilience are essential for the ‘ingest’ process, where live video data is sent to a server. The pivotal role of GSLB is to route traffic to the server that offers the best network quality relative to the user’s location. This not only improves the latency but also enhances the robustness and reliability of the streaming service.

Hosted GSLB Tailored for Streaming

Ceeblue’s GSLB is a specialized, cloud-based solution designed specifically for the streaming sector. Our system is integrated directly within our robust API ecosystem, ensuring that it is uniquely aligned with the needs and workflows of streaming platforms. Unlike generic SaaS-based GSLBs, Ceeblue’s offering is designed to cater specifically to the demands of streaming content, removing the need for additional provisioning, configuration, or maintenance.

Seamless Integration with Ceeblue APIs

The true power of Ceeblue’s GSLB lies in its deep integration with our existing API ecosystem. This integration means that developers working on streaming platforms don’t have to spend valuable time and resources developing a separate system to interface with our services. Instead, they can directly leverage our GSLB functionality through our APIs, ensuring a harmonious and efficient integration.

Zero Configuration and Streamlined Deployment

Ceeblue’s GSLB is designed for immediate activation and use, requiring no additional configuration from developers. It automatically identifies the most optimal regions for routing user traffic, which is particularly critical for maintaining high-quality streaming experiences. This approach significantly reduces the time and effort typically associated with setting up and managing a global server load balancer, enabling a quicker time to market for streaming services.

Enhanced Security and Simplified Operations

Ceeblue’s GSLB doesn’t just streamline traffic management; it also fortifies security at the network edge. Unwanted or unauthorized requests are intercepted before reaching the origin servers, ensuring that only legitimate traffic is processed. This feature is seamlessly integrated within our ecosystem, meaning developers can focus more on their core product rather than on implementing complex security measures.

Simplifying the Management of Ingest Endpoints

The Role of Persistent Ingest Endpoints in GSLB

The concept of a persistent ingest endpoint, initially sparked by a request by one of Ceeblue’s customers, is a key byproduct of GSLB implementation in streaming contexts. A persistent ingest endpoint refers to a stable, unchanging URL or IP address that streamers use to send their live video feeds. This stability is crucial for broadcasters who require a consistent and reliable connection for their streams.

How Persistent Ingest Endpoints Work in GSLB

In traditional streaming setups without GSLB, a streamer connects to a fixed server, and this connection does not change unless manually reconfigured. However, this approach can be problematic if the server faces issues like high traffic, outages, or maintenance. GSLB addresses these challenges by dynamically routing the stream to the optimal server without changing the endpoint the streamer connects to.

For example, consider a broadcaster in Paris using the persistent endpoint fhd1.in.cee.blue. Initially, their stream might be routed to a server in Paris. However, if this server becomes congested or experiences technical difficulties, the GSLB system can reroute the stream to a different server, say in Frankfurt, without requiring the broadcaster to change their streaming settings. The endpoint remains constant, but the underlying server handling the stream can change based on network conditions.

Benefits of Persistent Ingest Endpoints in GSLB Systems

  • Consistency: Broadcasters enjoy a consistent point of connection, simplifying their streaming setup and reducing the need for technical intervention.
  • Flexibility and Scalability: As the audience grows and network conditions vary, the GSLB system can scale and adapt by rerouting traffic to different servers.
  • Enhanced Reliability: By automatically switching to the best server based on real time network conditions, GSLB ensures higher uptime and reduces the impact of any single server’s failure.
  • Optimized Performance: Streamers benefit from reduced latency and improved stream quality as their content is served from the best available server.

Understanding Network Quality Factors and their Impact on Performance

At Ceeblue, when we refer to the term “Global” in the context of server balancing, we mean that our systems are distributed across the world in a global context. In a future post, we may delve deeper into the “Load” aspect of server balancing. For now, it’s essential to understand that the “Global” aspect of our server balancing strategy involves distributed systems and the importance of network quality between them.

We have defined four critical factors that play a key role in how we define network quality:

  • Latency: The time it takes for data to travel between two points in a network.
  • Location: The physical position of servers and users in relation to each other.
  • Congestion: The level of network traffic and the potential for bottlenecks.
  • Throughput: The amount of data that can be transmitted over a network in a given amount of time.

Although latency and location are commonly linked, our historical data reveals that they do not always have a direct correlation. This insight is crucial for GSLB strategies, where the assumption that closer geographic proximity always results in lower latency can lead to suboptimal server selection.

Regarding congestion and throughput, these are dynamic aspects of network performance that are inherently difficult to forecast. However, historical data can be a valuable asset in evaluating the peering quality between servers and users’ autonomous networks. By analyzing past performance and traffic patterns, GSLB systems can make more informed decisions, optimizing network management and enhancing user experience by intelligently routing traffic.

This approach allows our GSLB to not only consider proximity and latency but also to factor in the overall quality and reliability of network paths, leading to more efficient balancing and improved performance stability.

GSLB for Stream Ingestion based on Network Quality

A model for routing based on network quality, devised to determine the optimal Point of Presence (PoP) for stream ingestion.

The diagram shown in Figure 2 is a model for routing based on network quality, devised to determine the optimal Point of Presence (PoP) for stream ingestion.

When a stream is initiated to d6et.in.cee.blue, commonly referred to as a persistent ingest endpoint, the process begins with a DNS query. The DNS server resolves this query by providing the IP address of a suitable Edge Node at the Point of Presence (PoP) that offers the best network quality, taking into account factors like geographical proximity, the Autonomous System Number (AS Number), and latency. For example, the selected PoP could be nl-ams-1 if it offers the optimal network conditions.

The Ceeblue Global Traffic Director is integral to managing this routing strategy. When it receives a request, the Traffic Director evaluates the network quality of each PoP in relation to the user. Taking an Amsterdam-based user as an example, the Traffic Director would assess the network quality from Amsterdam to all PoPs in the network, considering their location, AS Number, and latency. If a PoP in Japan provides superior network quality, the request would be routed there. However, if Amsterdam’s local PoP scores higher on these metrics, the Traffic Director will route the traffic locally.

This approach to routing based on network quality is crucial because the internet’s performance is variable, influenced by ongoing changes in routing paths and network congestion. Therefore, a streamer who is optimally served by the US East PoP today might find a more suitable connection through the US West PoP tomorrow. By continuously analyzing and selecting the PoP with the highest network quality, the system ensures consistent and reliable streaming performance.

In conclusion, Global Server Load Balancing represents a sophisticated approach to managing network traffic, crucial for maintaining high-performance and reliable streaming services. Ceeblue’s tailored GSLB solution stands out by offering deep API integration, zero configuration requirements, and robust security features, all designed to meet the unique demands of streaming platforms. By leveraging persistent ingest endpoints and advanced network quality assessments, Ceeblue ensures that streamers can deliver high-quality content consistently and efficiently.

EXPLORE THE CEEBLUE WEBRTC CLIENT SDK

Source Code
github.com/CeeblueTV/webrtc-client

Prebuilt NPM Package
npmjs.com/package/@ceeblue/webrtc-client

TRY OUT THE WEBRTC VIDEO.JS PLUGIN

github.com/CeeblueTV/videojs-plugins

GET YOUR FREE CEEBLUE ACCOUNT

ceeblue.net/free-trial/

Keep up to date with the latest from Ceeblue

More To Explore

Live Video Transcoding and Streaming in Under a Second. Really.

Follow Us

OFFICES

Netherlands

Germany

Spain

©2021 Ceeblue B.V. All Rights Reserved. Privacy & Cookies Policy