Amazon announced its quarterly earnings last week. Since then, analysts have been reading a lot into how the retail giant's aggressive pricing war with Google has affected its overall revenues and profits.
Price cuts to Amazon Web Services are nothing new (Amazon has enacted more than 40 of them since 2006), but tying them so closely to their parent company's bottom line is novel. If anything, Amazon's latest figures show how AWS has become a mature and integral part of its business.
For now, the battle with Google et al is taking a modest toll. While revenue increased nearly one-quarter (year-over-year) to $19.3 billion, Amazon posted a $126 million loss, much larger than during the same period in 2013.
It's no time to panic for AWS users, though. Growth is still strong, and despite rising competition from Google and Microsoft, there are more options than ever for tweaking AWS for specific requirements.
Indeed, combining AWS with on-premises private cloud infrastructure is increasingly popular among enterprises keen to balance control and scalability. The hybrid cloud approach will also be a key way to stave off lock-in as a few large vendors come to lead the market. We'll look at hybrid uptake and similar trends in this week's Eucalyptus and cloud news roundup.
Weaker than Expected Amazon Earnings may be Explained by AWS Pricing Strategy
Writing for CloudPundit, Gartner's Lydia Leong put Amazon's earnings - and AWS pricing - into context. Essentially, this quarter's weak numbers should not be mistaken for a proxy either for declining interest in IaaS in general or for deterioration of AWS' enormous market share in particular.
Delivering services at near-cost isn't great for the bottom line, but in the long run it does several important things, such as:
- Driving off competitors that cannot compete on price or match the pace of innovation
- Widening the addressable market by removing the incentive to do everything on-premises
"The size of the price cut certainly had a negative impact on AWS's revenues this past quarter," explained Leong. "The price cut is likely larger than AWS would have done without pressure from Google. At the same time, the slight dip in revenue, versus the magnitude of the cuts, makes it clear that AWS is still growing at a staggeringly fast pace."
Leong also looked at the impact of both Google, Microsoft, IBM and others on AWS' position. She predicted continued dominance for AWS while noting that it finally has "credible competitors."
Amazon's Business as "Pain Management" Rather than IaaS
MongoDB's Matt Asay was also bullish about AWS, based on the company's success in other core operations like retail as well its differentiation from its competitors (Leong made a similar case in arguing that IaaS was not a commodity market).
It took Amazon six years to get to 1 trillion objects stored, but only 12 months to double that amount. Amazon CTO Werner Vogels expects a hockey stick-shaped growth trajectory.
Boosting Amazon's prospects is its focus on what Vogels calls "pain management for enterprises," rather than the standard term IaaS. In other words, it is addressing the traditional shortcomings of enterprise software inside the firewall.
Still, as Asay highlighted in the article for ReadWrite, the continual growth of AWS raises questions about data lock-in. Its dominance, or even the consolidation of the market around a handful of giant players, should make organizations think about how to keep options open through the use of open source software.
"What to a customer first looked like an exciting new piece of software - easy to try, no strings attached - soon infests the organization and isn't quite as easy to remove," stated Eucalyptus CEO Marten Mickos, according to Asay. "By using industry-standard open source software products, you reduce your lock-in down to an absolute minimum."
Mickos also posted an interesting chart to his Twitter account, laying out the three types of lock-in and how they can be avoided.
What are the Main Reasons for Private Cloud Failure?
If properly set up, private cloud is an excellent alternative to public infrastructure. High levels of security, control and performance can help companies save money in the long run while supporting even demanding applications. Plus, there's the opportunity to offload workloads to AWS as needed. But success in the private cloud is far from guaranteed. David Linthicum recently authored a piece for TechTarget explaining some of the common reasons behind substandard private cloud implementation:
- Too much repurposing of existing hardware and software: Private cloud operating systems often set a high bar for performance that legacy assets cannot meet.
- Overly loose management: A private cloud still requires diligence in areas such as provisioning and network access.
- Security oversight: Security is nominally a strength of private cloud, but organizations may get complacent. It's important to pay attention to the integrity of internal systems.
The Rising Popularity of Hybrid Integrations in the Enterprise
Hybrid cloud has picked up plenty of steam in the enterprise as a means of getting the best features of both public and private cloud. TechTarget associate editor Maxine Gaza looked at the Hybrid Cloud Customer Research conducted by Technology Business Research, which found that one-fifth of large enterprises had already set up two or more cloud-based systems.
Their reasons for doing so covered the usual bases, such as reducing total cost of ownership and making infrastructure more responsive to changing workloads. Roughly half of respondents stated that hybrid cloud had always been on their IT roadmap, while the remainder reported that it arose to meet new requirements or as a way to evolve existing products and services.
While plans are widespread, implementation can take time. Just as Linthicum highlighted the pitfalls of working with private cloud, analogous issues can set back hybrid setup, too.
"When customers purchase different clouds, they might not know upfront how difficult some of the integration will be," observed TBR senior analyst Jillian Mirandi, according to TechTarget.
"What is DevOps?" Two Years On
A few years back, O'Reilly Media vice president Mike Loukides wrote an influential blog post entitled "What is DevOps?" that outlined how development and operations had become siloed over the years. He recently penned an update that looks at ]DevOps more broadly as a way toward better corporate management](http://radar.oreilly.com/2014/06/revisiting-what-is-devops.html).
Loukides looked at the growing role of software in companies across all industries, which has raised the stakes for high-quality communication and teamwork. He also pointed out that DevOps is something that is enabled by technical tools; it is not a methodology defined by what specific solutions are used.
"It's always easy to think of DevOps (or of any software industry paradigm) in terms of the tools you use; in particular, it's very easy to think that if you use Chef or Puppet for automated configuration, Jenkins for continuous integration, and some cloud provider for on-demand server power, that you're doing DevOps," explained Loukides. "But DevOps isn't about tools; it's about culture, and it extends far beyond the cubicles of developers and operators.