|By Peter Lubbers, Frank Greco||
|March 11, 2010 10:15 AM EST||
Lately there has been a lot of buzz around HTML5 Web Sockets, which defines a full-duplex communication channel that operates through a single socket over the Web. HTML5 Web Sockets is not just another incremental enhancement to conventional HTTP communications; it represents a colossal advance, especially for real-time, event-driven web applications.
HTML5 Web Sockets provides such a dramatic improvement from the old, convoluted "hacks" that are used to simulate a full-duplex connection in a browser that it prompted Google's Ian Hickson - the HTML5 specification lead - to say:
"Reducing kilobytes of data to 2 bytes...and reducing latency from 150ms to 50ms is far more than marginal. In fact, these two factors alone are enough to make Web Sockets seriously interesting to Google."
Let's look at how HTML5 Web Sockets can offer such an incredibly dramatic reduction of unnecessary network traffic and latency by comparing it to conventional solutions.
Polling, Long-Polling, and Streaming - Headache 2.0
Normally when a browser visits a web page, an HTTP request is sent to the web server that hosts that page. The web server acknowledges this request and sends back the response. In many cases - for example, for stock prices, news reports, ticket sales, traffic patterns, medical device readings, and so on - the response could be stale by the time the browser renders the page. If you want to get the most up-to-date "real-time" information, you can constantly refresh that page manually, but that's obviously not a great solution.
With polling, the browser sends HTTP requests at regular intervals and immediately receives a response. This technique was the first attempt for the browser to deliver real-time information. Obviously, this is a good solution if the exact interval of message delivery is known, because you can synchronize the client request to occur only when information is available on the server. However, real-time data is often not that predictable, making unnecessary requests inevitable and, as a result, many connections are opened and closed needlessly in low-message-rate situations.
With long-polling, the browser sends a request to the server and the server keeps the request open for a set period. If a notification is received within that period, a response containing the message is sent to the client. If a notification is not received within the set time period, the server sends a response to terminate the open request. It is important to understand, however, that when you have a high message volume, long-polling does not provide any substantial performance improvements over traditional polling. In fact, it could be worse, because the long-polling might spin out of control into an unthrottled, continuous loop of immediate polls.
With streaming, the browser sends a complete request, but the server sends and maintains an open response that is continuously updated and kept open indefinitely (or for a set period of time). The response is then updated whenever a message is ready to be sent, but the server never signals to complete the response, thus keeping the connection open to deliver future messages. However, since streaming is still encapsulated in HTTP, intervening firewalls and proxy servers may choose to buffer the response, increasing the latency of the message delivery. Therefore, many streaming Comet solutions fall back to long-polling in case a buffering proxy server is detected. Alternatively, TLS (SSL) connections can be used to shield the response from being buffered, but in that case the setup and tear down of each connection taxes the available server resources more heavily.
Ultimately, all of these methods for providing real-time data involve HTTP request and response headers, which contain lots of additional, unnecessary header data and introduce latency. On top of that, full-duplex connectivity requires more than just the downstream connection from server to client. In an effort to simulate full-duplex communication over half-duplex HTTP, many of today's solutions use two connections: one for the downstream and one for the upstream. The maintenance and coordination of these two connections introduces significant overhead in terms of resource consumption and adds lots of complexity. Simply put, HTTP wasn't designed for real-time, full-duplex communication as you can see in Figure 1, which shows the complexities associated with building a Comet web application that displays real-time data from a back-end data source using a publish/subscribe model over half-duplex HTTP.
Figure 1: The complexity of Comet applications
It gets even worse when you try to scale out those Comet solutions to the masses. Simulating bi-directional browser communication over HTTP is error-prone and complex and all that complexity does not scale. Even though your end users might be enjoying something that looks like a real-time web application, this "real-time" experience has an outrageously high price tag. It's a price that you will pay in additional latency, unnecessary network traffic, and a drag on CPU performance.
HTML5 Web Sockets to the Rescue!
Defined in the Communications section of the HTML5 specification, HTML5 Web Sockets represent the next evolution of web communications - a full-duplex, bidirectional communications channel that operates through a single socket over the Web. HTML5 Web Sockets provides a true standard that you can use to build scalable, real-time web applications. In addition, since it provides a socket that is native to the browser, it eliminates many of the problems Comet solutions are prone to. Web Sockets remove the overhead and dramatically reduce complexity.
To establish a WebSocket connection, the client and server upgrade from the HTTP protocol to the WebSocket protocol during their initial handshake, as shown in the following example:
Example 1: The WebSocket handshake (browser request and server response)
GET /text HTTP/1.1\r\n
HTTP/1.1 101 WebSocket Protocol Handshake\r\n
Once established, WebSocket data frames can be sent back and forth between the client and the server in full-duplex mode. Both text and binary frames can be sent full-duplex, in either direction at the same time. The data is minimally framed with just two bytes. In the case of text frames, each frame starts with a 0x00 byte, ends with a 0xFF byte, and contains UTF-8 data in between. WebSocket text frames use a terminator, while binary frames use a length prefix.
The Showdown: Comet vs HTML5 Web Sockets
How dramatic is that reduction in unnecessary network traffic and latency? Let's compare a polling application with a WebSocket application.
For the polling example, I created a simple web application in which a web page requests real-time stock data from a RabbitMQ message broker using a traditional publish/subscribe model. It does this by polling a Java servlet that is hosted on a web server. The RabbitMQ message broker receives data from a fictitious stock price feed with continuously updating prices. The web page connects and subscribes to a specific stock channel (a topic on the message broker) and uses an XMLHttpRequest to poll for updates once per second. When updates are received, some calculations are performed and the stock data is shown in a table as shown in Figure 2.
Note: The back-end stock feed actually produces a lot of stock price updates per second, so using polling at one-second intervals is actually more prudent than using a Comet long-polling solution, which would result in a series of continuous polls. Polling effectively throttles the incoming updates here.
It all looks great, but a look under the hood reveals there are some serious issues with this application. For example, in Mozilla Firefox with Firebug (a Firefox add-on that allows you to debug web pages and monitor the time it takes to load pages and execute scripts), you can see that GET requests hammer the server at one-second intervals. Turning on Live HTTP Headers (another Firefox add-on that shows live HTTP header traffic) reveals the shocking amount of header overhead that is associated with each request. The following two examples show the HTTP header data for just a single request and response:
Example 2: HTTP request header
GET /PollingStock//PollingStock HTTP/1.1
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:22.214.171.124) Gecko/20091102 Firefox/3.5.5
Cookie: showInheritedConstant=false; showInheritedProtectedConstant=false; showInheritedProperty=false; showInheritedProtectedProperty=false; showInheritedMethod=false; showInheritedProtectedMethod=false; showInheritedEvent=false; showInheritedStyle=false; showInheritedEffect=false
Example 3: HTTP response header
HTTP/1.x 200 OK
Server: Sun Java System Application Server 9.1_02
Date: Sat, 07 Nov 2009 00:32:46 GMT
Just for fun, I counted all the characters. The total HTTP request and response header information overhead contains 871 bytes and that doesn't even include any data. Of course, this is just an example and you can have less than 871 bytes of header data, but I have also seen cases where the header data exceeded 2,000 bytes. In this example application, the data for a typical stock topic message is only about 20 characters long. As you can see, it's effectively drowned out by the excessive header information, which wasn't even required in the first place.
What happens when you deploy this application to a large number of users? Let's take a look at the network throughput for just the HTTP request and response header data associated with this polling application in three different use cases.
- Use case A: 1,000 clients polling every second:
Network throughput is (871 x 1,000) = 871,000 bytes = 6,968,000 bits per second (6.6 Mbps)
- Use case B: 10,000 clients polling every second:
Network throughput is (871 x 10,000) = 8,710,000 bytes = 69,680,000 bits per second (66 Mbps)
- Use case C: 100,000 clients polling every 1 second:
Network throughput is (871 x 100,000) = 87,100,000 bytes = 696,800,000 bits per second (665 Mbps)
That's an enormous amount of unnecessary network throughput. If only we could just get the essential data over the wire. Well, guess what? You can with HTML5 Web Sockets. I rebuilt the application to use HTML5 Web Sockets, adding an event handler to the web page to asynchronously listen for stock update messages from the message broker (check out the many how-tos and tutorials on www.tech.kaazing.com/documentation for more information on how to build a WebSocket application). Each of these messages is a WebSocket frame that has just two bytes of overhead (instead of 871). Take a look at how that affects the network throughput overhead in our three use cases.
- Use case A: 1,000 clients receive 1 message per second:
Network throughput is (2 x 1,000) = 2,000 bytes = 16,000 bits per second (0.015 Mbps)
- Use case B: 10,000 clients receive 1 message per second:
Network throughput is (2 x 10,000) = 20,000 bytes = 160,000 bits per second (0.153 Mbps)
- Use case C: 100,000 clients receive 1 message per second:
Network throughput is (2 x 100,000) = 200,000 bytes = 1,600,000 bits per second (1.526 Mbps)
As you can see in Figure 3, HTML5 Web Sockets provide a dramatic reduction of unnecessary network traffic compared to the polling solution.
Figure 3: Comparison of the unnecessary network throughput overhead between the polling and the WebSocket applications
What about the reduction in latency? Take a look at Figure 4. In the top half, you can see the latency of the half-duplex polling solution. If we assume, for this example, that it takes 50 milliseconds for a message to travel from the server to the browser, then the polling application introduces a lot of extra latency, because a new request has to be sent to the server when the response is complete. This new request takes another 50ms and during this time the server cannot send any messages to the browser, resulting in additional server memory consumption.
In the bottom half of the figure, you see the reduction in latency provided by the WebSocket solution. Once the connection is upgraded to WebSocket, messages can flow from the server to the browser the moment they arrive. It still takes 50 ms for messages to travel from the server to the browser, but the WebSocket connection remains open so there is no need to send another request to the server.
Figure 4: Latency comparison between the polling and WebSocket applications
HTML5 Web Sockets and the Kaazing WebSocket Gateway
Today, only Google's Chrome browser supports HTML5 Web Sockets natively, but other browsers will soon follow. To work around that limitation, however, Kaazing WebSocket Gateway provides complete WebSocket emulation for all the older browsers (I.E. 5.5+, Firefox 1.5+, Safari 3.0+, and Opera 9.5+), so you can start using the HTML5 WebSocket APIs today.
WebSocket is great, but what you can do once you have a full-duplex socket connection available in your browser is even greater. To leverage the full power of HTML5 Web Sockets, Kaazing provides a ByteSocket library for binary communication and higher-level libraries for protocols like Stomp, AMQP, XMPP, IRC and more, built on top of WebSocket.
For example, if you use a high-level library for protocols such as Stomp or AMQP (supplied with the Kaazing Gateway), you can talk directly to back-end message brokers like the RabbitMQ broker described in the example. By having a direct connection to services, there is no need for additional application server logic to translate those bi-directional, full-duplex TCP backend protocols to uni-directional, half-duplex HTTP connections; the browser can simply understand these protocols natively.
Figure 5: Kaazing WebSocket Gateway extends TCP-based messaging to the browser with ultra high performance
HTML5 Web Sockets provides an enormous step forward in the scalability of the real-time web. As you have seen in this article, HTML5 Web Sockets can provide a 500:1 or - depending on the size of the HTTP headers - even a 1000:1 reduction in unnecessary HTTP header traffic and 3:1 reduction in latency. That is not just an incremental improvement; that is a revolutionary jump - a quantum leap.
Kaazing WebSocket Gateway makes HTML5 WebSocket code work in all the browsers today, while providing additional protocol libraries that allow you to harness the full power of the full-duplex socket connection that HTML5 Web Sockets provides and communicate directly to back-end services. For more information about Kaazing WebSocket Gateway, visit www.kaazing.com and the Kaazing technology network at tech.kaazing.com.
- HTML5 Web Sockets Specification: http://dev.w3.org/html5/websockets/
- WebSockets.org: http://www.websockets.org
- Kaazing corporation: http://www.kaazing.com
- Download the Kaazing WebSocket Gateway: http://www.kaazing.com/download
- Skills Matter HTML5 Communication Training Courses: http://skillsmatter.com/expert-profile/java-jee/peter-lubbers
- Book: Pro HTML5 Programming: Powerful APIs for Richer Internet Application Development (Link)
|PeterLubbers 03/15/10 03:55:00 PM EDT|
An additional problem with streaming is that you can get unexpected results if there are intermediaries such as buffering proxy servers in the way. Streaming solutions therefore often fall back to long-polling if they detect such intermediaries (if they are smart enough to detect them at all). With WebSocket and WebSocket secure you have a better chance of traversing proxy servers and firewalls.
|tbayes 03/11/10 10:11:00 AM EST|
Hi Peter and Frank,
Thank you for this clear and well-written article, which does much to effectively explain the benefits of the WebSocket protocol.
If I may point out one minor correction: In the WebSocket section of Example 3, Use case B and C should show overhead rates of 0.153Mbps and 1.526Mbps respectively, as opposed to 0.153Kbps and 1.526Kbps.
Perhaps a more natural comparison would be between Web Sockets and Comet-streaming (via, for example, the "Forever IFrame" technique), as I feel it a little unsporting to compare Web Sockets directly to polling or long-polling! At around 20 bytes per message, the overhead associated with clients receiving messages via Comet-style streaming is still more than WebSocket's 2 bytes per message, but is still one to two orders of magnitude less than the overhead associated with polling techniques. Similarly, latency whether receiving messages using Comet-streaming or WebSocket is the same. Of course, clients sending messages will gain the advantages you describe only with Web Sockets, as all other approaches do indeed require new HTTP requests.
I completely agree, however, that WebSocket is a definite improvement over the Comet IFrame-approaches, which have an air of string and glue about them.
I hope that we will see WebSocket support in all production mainstream browsers very soon.
What does “big enough” mean? It’s sometimes useful to argue by reductio ad absurdum. Hello, world doesn’t need to be broken down into smaller services. At the other extreme, building a monolithic enterprise resource planning (ERP) system is just asking for trouble: it’s too big, and it needs to be decomposed.
Aug. 29, 2015 10:00 AM EDT Reads: 321
Early in my DevOps Journey, I was introduced to a book of great significance circulating within the Web Operations industry titled The Phoenix Project. (You can read our review of Gene’s book, if interested.) Written as a novel and loosely based on many of the same principles explored in The Goal, this book has been read and referenced by many who have adopted DevOps into their continuous improvement and software delivery processes around the world. As I began planning my travel schedule last...
Aug. 29, 2015 10:00 AM EDT Reads: 499
Docker containerization is increasingly being used in production environments. How can these environments best be monitored? Monitoring Docker containers as if they are lightweight virtual machines (i.e., monitoring the host from within the container), with all the common metrics that can be captured from an operating system, is an insufficient approach. Docker containers can’t be treated as lightweight virtual machines; they must be treated as what they are: isolated processes running on hosts....
Aug. 29, 2015 10:00 AM EDT
SYS-CON Events announced today that G2G3 will exhibit at SYS-CON's @DevOpsSummit Silicon Valley, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. Based on a collective appreciation for user experience, design, and technology, G2G3 is uniquely qualified and motivated to redefine how organizations and people engage in an increasingly digital world.
Aug. 29, 2015 09:30 AM EDT Reads: 426
SYS-CON Events announced today that HPM Networks will exhibit at the 17th International Cloud Expo®, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. For 20 years, HPM Networks has been integrating technology solutions that solve complex business challenges. HPM Networks has designed solutions for both SMB and enterprise customers throughout the San Francisco Bay Area.
Aug. 29, 2015 09:30 AM EDT Reads: 836
Introducing Containers & Microservices Bootcamp at @CloudExpo Silicon Valley | #Containers #Microservices
SYS-CON Events announced today the Containers & Microservices Bootcamp, being held November 3-4, 2015, in conjunction with 17th Cloud Expo, @ThingsExpo, and @DevOpsSummit at the Santa Clara Convention Center in Santa Clara, CA. This is your chance to get started with the latest technology in the industry. Combined with real-world scenarios and use cases, the Containers and Microservices Bootcamp, led by Janakiram MSV, a Microsoft Regional Director, will include presentations as well as hands-on...
Aug. 29, 2015 09:15 AM EDT Reads: 151
The pricing of tools or licenses for log aggregation can have a significant effect on organizational culture and the collaboration between Dev and Ops teams. Modern tools for log aggregation (of which Logentries is one example) can be hugely enabling for DevOps approaches to building and operating business-critical software systems. However, the pricing of an aggregated logging solution can affect the adoption of modern logging techniques, as well as organizational capabilities and cross-team ...
Aug. 29, 2015 08:30 AM EDT Reads: 360
In today's digital world, change is the one constant. Disruptive innovations like cloud, mobility, social media, and the Internet of Things have reshaped the market and set new standards in customer expectations. To remain competitive, businesses must tap the potential of emerging technologies and markets through the rapid release of new products and services. However, the rigid and siloed structures of traditional IT platforms and processes are slowing them down – resulting in lengthy delivery ...
Aug. 29, 2015 07:45 AM EDT Reads: 549
Skeuomorphism usually means retaining existing design cues in something new that doesn’t actually need them. However, the concept of skeuomorphism can be thought of as relating more broadly to applying existing patterns to new technologies that, in fact, cry out for new approaches. In his session at DevOps Summit, Gordon Haff, Senior Cloud Strategy Marketing and Evangelism Manager at Red Hat, discussed why containers should be paired with new architectural practices such as microservices rathe...
Aug. 29, 2015 06:00 AM EDT Reads: 370
Any Ops team trying to support a company in today’s cloud-connected world knows that a new way of thinking is required – one just as dramatic than the shift from Ops to DevOps. The diversity of modern operations requires teams to focus their impact on breadth vs. depth. In his session at DevOps Summit, Adam Serediuk, Director of Operations at xMatters, Inc., will discuss the strategic requirements of evolving from Ops to DevOps, and why modern Operations has begun leveraging the “NoOps” approa...
Aug. 29, 2015 05:30 AM EDT Reads: 342
Puppet Labs has announced the next major update to its flagship product: Puppet Enterprise 2015.2. This release includes new features providing DevOps teams with clarity, simplicity and additional management capabilities, including an all-new user interface, an interactive graph for visualizing infrastructure code, a new unified agent and broader infrastructure support.
Aug. 29, 2015 03:00 AM EDT Reads: 473
DevOps has traditionally played important roles in development and IT operations, but the practice is quickly becoming core to other business functions such as customer success, business intelligence, and marketing analytics. Modern marketers today are driven by data and rely on many different analytics tools. They need DevOps engineers in general and server log data specifically to do their jobs well. Here’s why: Server log files contain the only data that is completely full and accurate in th...
Aug. 29, 2015 12:15 AM EDT Reads: 331
SYS-CON Events announced today that Pythian, a global IT services company specializing in helping companies leverage disruptive technologies to optimize revenue-generating systems, has been named “Bronze Sponsor” of SYS-CON's 17th Cloud Expo, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. Founded in 1997, Pythian is a global IT services company that helps companies compete by adopting disruptive technologies such as cloud, Big Data, advance...
Aug. 28, 2015 12:00 PM EDT Reads: 203
It’s been proven time and time again that in tech, diversity drives greater innovation, better team productivity and greater profits and market share. So what can we do in our DevOps teams to embrace diversity and help transform the culture of development and operations into a true “DevOps” team? In her session at DevOps Summit, Stefana Muller, Director, Product Management – Continuous Delivery at CA Technologies, answered that question citing examples, showing how to create opportunities for ...
Aug. 28, 2015 12:00 PM EDT Reads: 446
Several years ago, I was a developer in a travel reservation aggregator. Our mission was to pull flight and hotel data from a bunch of cryptic reservation platforms, and provide it to other companies via an API library - for a fee. That was before companies like Expedia standardized such things. We started with simple methods like getFlightLeg() or addPassengerName(), each performing a small, well-understood function. But our customers wanted bigger, more encompassing services that would "do ...
Aug. 28, 2015 11:00 AM EDT Reads: 177
Culture is the most important ingredient of DevOps. The challenge for most organizations is defining and communicating a vision of beneficial DevOps culture for their organizations, and then facilitating the changes needed to achieve that. Often this comes down to an ability to provide true leadership. As a CIO, are your direct reports IT managers or are they IT leaders? The hard truth is that many IT managers have risen through the ranks based on their technical skills, not their leadership ab...
Aug. 28, 2015 10:00 AM EDT Reads: 276
SYS-CON Events announced today that DataClear Inc. will exhibit at the 17th International Cloud Expo®, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. The DataClear ‘BlackBox’ is the only solution that moves your PC, browsing and data out of the United States and away from prying (and spying) eyes. Its solution automatically builds you a clean, on-demand, virus free, new virtual cloud based PC outside of the United States, and wipes it clean...
Aug. 28, 2015 09:45 AM EDT Reads: 341
Whether you like it or not, DevOps is on track for a remarkable alliance with security. The SEC didn’t approve the merger. And your boss hasn’t heard anything about it. Yet, this unruly triumvirate will soon dominate and deliver DevSecOps faster, cheaper, better, and on an unprecedented scale. In his session at DevOps Summit, Frank Bunger, VP of Customer Success at ScriptRock, will discuss how this cathartic moment will propel the DevOps movement from such stuff as dreams are made on to a prac...
Aug. 28, 2015 09:45 AM EDT Reads: 184
In his session at 17th Cloud Expo, Ernest Mueller, Product Manager at Idera, will explain the best practices and lessons learned for tracking and optimizing costs while delivering a cloud-hosted service. He will describe a DevOps approach where the applications and systems work together to track usage, model costs in a granular fashion, and make smart decisions at runtime to minimize costs. The trickier parts covered include triggering off the right metrics; balancing resilience and redundancy ...
Aug. 28, 2015 09:30 AM EDT Reads: 129
The Microservices architectural pattern promises increased DevOps agility and can help enable continuous delivery of software. This session is for developers who are transforming existing applications to cloud-native applications, or creating new microservices style applications. In his session at DevOps Summit, Jim Bugwadia, CEO of Nirmata, will introduce best practices, patterns, challenges, and solutions for the development and operations of microservices style applications. He will discuss ...
Aug. 27, 2015 02:15 PM EDT Reads: 505