Amazon Web Services often gets criticized as a platform that doesn't necessarily scale for the enterprise. So at re:Invent, the second annual AWS conference, Amazon made a series of announcements aimed squarely at dispelling these concerns.
Virtualization and Cloud Advisor
By CIO Staff, CIO
Amazon Web Services’ second re:invent conference was a big, noisy, sprawling extravaganza spread across the multiple floors and endless hallways of the Las Vegas Venetian Convention Center. If a theme could be distilled from the conference, you could say it’s “Yeah, we do enterprise.”
Given that AWS has traditionally been characterized as a good resource for developers, startups and the SMB market, but not really ready to support “the enterprise,” the message that AWS is ready to meet the requirements of enterprise customers shouldn’t be a surprise. For the past couple years, AWS has relentlessly focused on enterprise needs. This year’s Re:invent is where all that effort came together.
Here are the seven most striking aspects of re:invent as they relate to an enterprise market.
1. New Customers Validating AWS Business Model
The CIO of News International explains that the organization is going from 40 data centers to 6 and plans to shift 75 percent of its computing to AWS. People pooh-pooh AWS users such as Netflix, Dropbox and Pinterest, saying they’re nothing like standard enterprises. The same can’t be said about News International, a representative, although not typical, enterprise
2. AWS Adoption Growing at Remarkable Rates
Andy Jassy, the head of AWS, shared the fact that RedShift, the data warehousing service announced at last year’s re:invent, is the fastest-growing service ever launched by AWS. When you consider the size and scale of some of the other AWS services — such as Simple Storage Service (S3), which manages more than 2 trillion objects — that’s a pretty remarkable statement.
Another stat Jassy mentioned: AWS Marketplace, in which AWS partners make their services available within AWS, has grown more than 700 percent this year. The services are pre-integrated to work with AWS and licensed such that customers can begin using them in their applications without needing to strike a separate deal with the provider.
This makes it much easier to get going with the partner products. It also makes AWS a more attractive application development environment, because many of the complementary products and services you might need to build an application are already integrated and available. People commonly underestimate the importance of ecosystem, but the blistering growth of the marketplace shows that AWS customers are avidly adopting AWS partner products.
3. AWS WorkSpaces Making VDI Easy
Re:invent is striking as to how casually Amazon drops announcements of new products. Things that would be showcased as the biggest event of another vendor’s conference are instead announced, described and moved on from. So it was with AWS WorkSpaces, a virtual desktop offering.
AWS will offer virtual desktops as a service priced well below current offerings, at $50 to $75 per month. VDI is exactly the boring type of enterprise service that keeps vendors churning out pleasing quarterly reports; there are hundreds of millions of desktops out in the world, and enterprises are desperate to find a better way to manage these powerful yet finicky devices. The immediate attractiveness of the AWS service could be seen in the abrupt drop in Citrix Systems’ stock at the moment of announcement.
If I had to guess, I’d say this will be a slowly adopted, but ultimately very large, AWS service. Enabling enterprises to get out of the on-site, break-fix PC business is bound to be enticing to companies both large and small.
4. AWS AppStream Enabling Mobile
It’s a cliché that mobile is the topic of the times, but it’s clear that enterprise IT groups are under tremendous pressure to deliver more mobile apps. Unfortunately, mobile applications come with many complexities, not least of which is how to move data back and forth between device and data center. This is only exacerbated by the explosion in the use of video in mobile apps.
Another Jassy announcement, AppStream, lets mobile apps offload computing and graphics processing to AWS and just stream the results out to a device. This can let lower-spec devices implement rich applications. It just might be the thing for IT organizations challenged to deliver applications to the rich stew of devices present in a BYOD world.
5. AWS Supporting High-computation Workloads
Another criticism of AWS, and a reason cited as to its enterprise-unworthiness, relates to its virtual machine capabilities. People criticize AWS for not providing the high-performance computing that’s characterized as necessary for enterprise workloads. During his keynote, AWS CTO Werner Vogels announced a new AWS instance family, the C3 type, which is based on Intel Ivy Bridge chips and uses solid state drive (SSD) storage for fast I/O. Enterprise workload support, done.
6. AWS Supporting Enterprise Compliance
Another oft-cited AWS enterprise shortcoming is the need for security, compliance and audit. Over the past couple of years, AWS has achieved just about every security certification under the sun, often pursued as part of its push into the government market.
The latest announcement, CloudTrail, addresses audit requirements. It tracks and records all AWS API calls and stores them in S3, where users can examine and parse the logs to see who made what API calls. This is relevant because all AWS activity is executed via API calls — so CloudTrail means that customers can have access to everything that anyone in the organization did in AWS.
7. AWS Transforming Disaster Recovery Market
Vogels dropped another titbit: Cross-region RDS read replicas. That sounds like jargon, but it carries powerful implications. RDS is AWS’ managed relational database service. It already has cross-availability zone replication, by which customers can protect their applications from AWS outages in one zone. (A zone may be conceptually considered as a single data center; availability zone replication insulates users from a data center outage.)
The extension of read replicas across regions is a big deal. Now customers can protect themselves against nearly any conceivable AWS outage, even one that somehow managed to knock an entire region off the air. This will transform the DR market because it makes it easy to keep replica systems available and ready to take over processing at a moment’s notice.
Traditionally, disaster recovery is difficult for two reasons: Having sufficient computing capacity available in a secondary location and getting application data to the secondary location when an outage occurs. AWS EC2 makes the first difficulty vanish — it’s now trivial to have identical VM images located in a backup region.
Getting application data to the second location, however, has been a challenge. With a read replica, this issue is rendered trivial as well. Getting a read replica up and running requires nothing more than a declaration at time of RDS resource creation (or subsequently declaring the RDS resource as carrying a read replica). With RDW cross-region read replicas, AWS has now addressed the two challenges of disaster recovery and made DR accessible to everyone.
In a presentation, AWS data center architect James Hamilton described how Amazon is buying fiber for its service. I’ve previously written about the unique hardware environment of AWS, speculating that this fiber is used (in part) to reduce latency in the cross-region read replicas. DR providers traditionally thrive by fitting expensive backup data centers with unused hardware that sits around, just waiting for someone to have a disaster. Using these facilities requires forwarding archive tapes to the location, installing application software, restoring backups and gradually bringing applications back online.
The Amazon announcement obviates that approach and makes restoring applications simpler — and much faster — with no need for tape archives and much less lost data as well. If I were a DR provider, I’d be shaking in my boots. This announcement carries the potential to disrupt a critical, multibillion dollar market — and, as I said, it was just one announcement among many.
Over the past few years, Amazon has steadily chipped away at the objections raised as to why enterprises can’t use AWS. The breadth and depth of enterprise-focused announcements at re:invent 2013 makes it clear that Amazon has enterprise users in its crosshairs as it builds out the service. The number of reasons enterprises have to avoid using AWS is steadily decreasing, and the stigma for an enterprise using AWS has nearly disappeared. If you want to see what the future of enterprise computing looks like, be sure to get to re:invent 2014.
Bernard Golden is senior director of Cloud Computing Enterprise Solutions group at Dell. Prior to that, he was vice president of Enterprise Solutions for Enstratius Networks, a cloud management software company, which Dell acquired in May 2013. He is the author of three books on virtualization and cloud computing, including Virtualization for Dummies. Follow Bernard Golden on Twitter @bernardgolden.