The importance of Riverbed’s layer-7 optimizations

Last week at a conference, I was approached by a prospective customer who was also considering an alternative offering from a Riverbed competitor.   While this competitive WAN optimization product lacked a number of layer-7 optimization capabilities available in the Riverbed solution (among other things), my questioner was wondering if its significantly-cheaper price made up for its lack of features and capabilities.

More specifically, the prospective customer had questions about the true value of Riverbed's layer-7 optimizations for Exchange.  Since his employees were all using Outlook Cache Mode, he reasoned that he didn't need lightning-fast transfer performance for email–his end-users wouldn't notice if emails took a little longer to show up in their Outlook client.  He felt that his primary requirement was for the compression and data reduction, since his branch offices all had smaller WAN links and were experiencing network congestion problems.  As far as he was concerned, he just needed a product that could deliver more virtual bandwidth, and he was under the impression that most WAN optimization vendors–including the Riverbed competitor that he was considering–would likely meet his requirement.

My response was to point out that Riverbed's layer-7 optimizations go far beyond just addressing latency and protocol chattiness issues–they also address application-specific data encoding processes that are performed by many applications.  In the case of Microsoft Exchange, all email attachments are encoded in a MAPI-specific proprietary data format.  At the byte-level, these MAPI attachments appear differently than if that same file had been transferred through another protocol such as FTP, CIFS or HTTP (i.e., Sharepoint).  Because Riverbed's layer-7 optimization feature recognizes the application-specific data encoding process, it is able to un-do the encoding prior to applying its data reduction algorithms.  When using Riverbed, files and other common data that had previously been sent over the network using other applications and protocols will be optimized through "warm" transfers, which greatly boosts overall data reduction results.  We refer to this benefit as "cross-protocol" optimization.

The data reduction benefits I am talking about are clearly illustrated in this screenshot submitted by a Riverbed user (click here) for the Performance Hall of Fame.  This customer notes that among other things, they replicate their Exchange DB using iSCSI through their Equalogic iSCSI SAN.  But note that there is also MAPI optimization taking place between these two sites, which means that the same files are actually traversing the WAN link multiple times–through Exchange/MAPI, as well as through iSCSI for the DR replication.

A competitive product lacking Riverbed's layer-7 optimization capabilities for MAPI would be confused by the encoding process that Exchange performs on all email attachments.  Such a product would think there is no data commonality between the MAPI and iSCSI sessions, as well as between the MAPI and CIFS sessions.  In addition, the competitive product would also be affected by minor changes to file attachments, such as when an employee edits a few lines of a Word document that he/she just received, and then emails that slightly-modified file back to the originator.  The problem here is that even minor changes to a small portion of a file will be rippled through entire length of the file attachment by the Exchange attachment encoding process.  The competitive product will think they are two different and unrelated files with no common data, when in fact they are actually the same file with a few minor changes.

Finally, it's important to note that Exchange is far from being the only application that performs application-specific data encoding.  Other applications that perform similar actions include Lotus Notes, Oracle E-Business Suite (both JRE and JInitiator), and Citrix ICA.  In addition, some applications such as SSL and encrypted MAPI send cryptographically encrypted data.  These application traffic types are well represented in most enterprise networks, including that of the Riverbed user providing the screenshot.  Riverbed's layer-7 optimizations for these listed applications deliver not only lightning-fast performance by addressing layer-7 protocol chattiness issues, but also superior overall data reduction by recognizing the layer-7 data encoding and/or encryption performed on the data itself by the application. 

Just What is Social Media Marketing?

Normally I am not the kind of person who hangs comic strips on his cubicle walls, but when Dilbert discusses your exact job, you kind of have to.  And since I don't actually have a cubicle (I work from home), this blog will have to act as my cubicle wall.  

I thought this might be a nice opportunity to talk about Riverbed's Social Media presence.  (Fortunately, Riverbed is enlightened enough to permit its Social Media people to use Twitter and Facebook from work!)

• First and foremost, of course, is this blog.  If you're here, then you've already found it.  If you're not here, then it opens up all kinds of metaphysical questions that I'd rather not spend a lot of time thinking about.
• Second is our Community Site where hundreds of Riverbed users exchange questions, answers, advice, and other general comments about their Riverbed products and implementations.  Right now, on the Community Site, our Performance Hall of Fame is under way, where users have the chance to win Flip Video Cameras by posting graphs from their real Riverbed implementations for everyone to see.  Everyone who enters wins a Special Edition Riverbed T-shirt.  If you haven't seen it yet, be sure to come by and check it out!
• Third is our Twitter feed, where you can find all sorts of announcements about what we're doing across the company as well as across our Social Media.  Please be sure to follow us if you aren't already.
• Fourth is our Facebook page.  Please be sure to Like us.
• We've also got a LinkedIn Group where there are some animated discussions and Q&As about various aspects about Riverbed, our products, and many of the things we do.
• We make a bunch of videos, and they can all be found together on our YouTube channel.  There are customer testimonials, Chalk Talks (brief videos that explain one aspect of our products or another), training videos, and lots of other things.  You can find them all in one place on our YouTube channel.  Please be sure to friend us on YouTube.
• And we're also on Google Buzz.

Please be sure to join us in our various Social Media incarnations.  We'll be watching for you.

Network World’s IT Roadmap show comes to Dallas

 

Gonna be in Dallas on September 14?

Going to the IT Roadmap Conference at the Dallas Convention Center?

Make sure you drop by the Riverbed booth on the Exposition Floor.  It'll be your chance to learn all about Riverbed Cascade, our product for end-to-end network management, and to see our famous Jack Demo.

While you're there, be sure to say hi to Lindsay and Brian, who'll be staffing the booth, and tell them you saw it on the Riverbed Blog!

Riverbed’s optimizations for Citrix ICA and other real-time traffic types

There has been some confusion lately regarding Riverbed's capabilities to optimize thin-client and other real-time traffic types, so in this blog I'd like to summarize all of the different ways that Riverbed enhances the performance of Citrix ICA, RDP, VoIP, and other similar traffic types.

Numerous Riverbed customers are using Riverbed to deliver significant bandwidth reduction and performance improvements for Citrix ICA and RDP traffic.  Typical data reduction results range from 20% to 60%, while response time improvements vary from 10% to 40%.  Actual results depend on various environmental factors; better data reduction results are generally achieved when a greater number of thin clients access data from a given remote site.  Response time improvements are generally more noticeable when there is pre-existing network congestion and performance issues prior to enabling the Riverbed optimizations.

Riverbed data deduplication capabilities are particularly effective when applied to thin client and other traffic types characterized by small packet sizes because of the granularity of the Steelhead algorithms.  On average, Riverbed is able to identify byte-level commonalities of only 100 bytes.  In contrast, many of Riverbed's competitors use algorithms that can only identify byte-level commonalities of 4KB to 32KB–a huge difference.  In addition, Riverbed Steelheads are able to implement their algorithms entirely in memory when applied to real-time traffic types such as RDP and Citrix ICA, in order to minimize processing latency and avoid adding jitter and latency to the real-time traffic.  Most of Riverbed's competitors lack this capability.

For Citrix ICA traffic in particular, Riverbed's optimizations are specialized and powerful.  The Riverbed Steelheads are actually able to communicate with the Citrix XenApp server's API's and disable its default compression on a per-flow basis.  The Steelheads can then apply their algorithms on the original uncompressed Citrix ICA data, allowing the Citrix connection to receive the full benefit of Riverbed's byte-level data reduction technology.  Generally, most competitive vendors require that compression settings be globally disabled on the Citrix server.  This has the drawback that all Citrix users–those local in the LAN as well as those at remote sites who do not have WAN optimization devices–receive uncompressed Citrix ICA traffic, which makes those users more vulnerable to bandwidth constraints and network congestion issues.  With Riverbed, only those flows being optimized by the Steelhead appliances will have their default compression disabled.

Riverbed Steelheads also have the ability to decode the ICA Priority Packet Tagging that identifies the virtual channel from which each Citrix ICA packet originated.  As part of this capability, Riverbed specifically developed a packet-order queuing discipline that respects the ordering of ICA packets within a flow, even when different packets from a given flow are classified by Citrix into different ICA virtual channels.  This allows the Steelhead to deliver very granular Quality of Service (QoS) enforcement based on the virtual channel in which the ICA data is transmitted.  Most importantly, this feature prevents any possibility of out-of-order packet delivery as a result of Riverbed's QoS enforcement; out-of-order packet delivery would cause significant degradation in performance and responsiveness for the Citrix ICA user.  Riverbed's packet-order queuing capability is patent-pending, and not available from any other WAN optimization vendor.

With regard to optimizing VoIP traffic, here again Riverbed has specialized capabilities.  Since the VoIP data is already efficiently compressed by the voice codec, it therefore is not possible to achieve further compression of the VoIP data (you should be wary of any vendor that tells you otherwise).  However, it is possible to apply very granular QoS enforcement for VoIP traffic to ensure the voice call remains unaffected by network congestion that may occur in the WAN.  Toward this end, Riverbed's Hierarchical Fair Service Curve (HFSC) scheduling mechanisms in its QoS enforcement capabilities can guarantee high-priority forwarding service for any type of traffic, including traffic types such as VoIP that consume a small amount of bandwidth. Traditional systems that use Packet Fair Queuing (PFQ) mechanisms, including those used by Riverbed's competitors, don't work as well in these settings.  The problem with PFQ-based systems is that low-latency forwarding service is only provided to traffic classes that have been allocated large amounts of bandwidth.  As a result, PFQ-based systems do not perform as well when handling high-priority traffic that consume small amounts of bandwidth, including traffic types such as VoIP, Citrix ICA, RDP, telnet, etc.

Here is a link to a real Riverbed customer who achieved 61% data reduction for Citrix ICA, saving 14GB of Citrix data per month that didn't have to be sent over the WAN. 

The Riverbed Performance Hall of Fame is back!

Last year, Riverbed ran our first-ever Performance Hall of Fame.  Over 130 of our community.riverbed.com readers submitted performance graphs and results for everyone to see.  Everybody who entered got a special edition Riverbed t-shirt, and was entered in a contest to win a cool prize (last year it was an iPod).    

Did you miss it?  

Don't worry… it's not too late.  We are running a whole new Performance Hall of Fame between now and the end of 2010.  Upload your Riverbed performance graphs depicting about a week's worth of amazing performance, write up a paragraph or two describing what you've done, give it a pithy title, and upload it.  (If you assign port names, that'd be even better!)

Just like last year, everyone who enters gets a t-shirt.  (Make sure your mailing address is correct in your registration so we can find you!)  

Then, vote for the best posts using the Kudos feature (yes, you can vote for yourself!), and we'll give prizes to the people who get the most Kudos.

The big change in the contest this year is that the prize is a Flip Video Camera.  We'll be awarding one per month between now and the end of the year.

If you like reading fine print, then by all means, click here!

And let's have some fun with this!

Good luck to everyone who enters.  And thanks!

Make those VM backups even FASTER!

Riverbed recently announced that Pancetera has joined the Riverbed Technology Alliance.  

Our friends over at the Pancetera Blog have blogged a couple of times (one and two) about how combining their Pancetera Unite product with Riverbed Steelhead Appliances can deliver even better performance for Virtual Machine and traditional backups across the WAN than either product can alone.  

My favorite quote from their blog is as follows:

A dedicated OC-3 line of 155 Mbps with Pancetera and Riverbed Steelhead will perform better than an OC-12 of 622 Mbps line at a quarter of the cost. Savings of $500,000+ per year provide a near instant ROI.