AWS has a lot to offer when it comes to data management. But how does it stack up against Google and Azure database...
AWS generally leads the public cloud pack with its breadth of services, while Google leads in performance and a performance-price ratio, according to Henri Binsztok, chief innovation officer at Wallix Group, a cybersecurity company based in Paris. Binsztok bases performance-price info on RightScale and The HFT Guy studies, as well as data from Google.
AWS provides no fewer than 15 storage and database services, plus four data migration services and nine analytics services, according to Binsztok. This is more than Google, which provides six storage and database services, plus another eight big data services. Meanwhile, Microsoft Azure provides 18 storage and database services and another eight data and analytics services.
"The key takeaway is that all three platforms provide a high number of data services," though they all vary somewhat in their offerings, Binsztok said. And those services divide qualitatively among storage, database and data services that encompass migration, big data or analytics.
Beyond the numbers
But number of available services only tells part of the story. Amazon Relational Database Service now comprises MySQL, PostgreSQL, Oracle, SQL Server and MariaDB, while the Microsoft platform counts each as separate services. Google, like AWS, counts its database service as one but does not provide the same maturity as AWS, Binsztok said. For example, Google's Cloud SQL for PostgreSQL is still in beta.
Google takes a platform-as-a-service approach to its data management offerings, and its application engineering heritage includes Bigtable, its NoSQL big data database service. "They pretty much invented that; there is nothing like that in the industry," said Brian Hopkins, analyst at Forrester Research. "You can shove as much data into it as you want."
But Google lacks real strength in metadata. That's why you have to be careful what you put into Google Cloud Platform, because getting data back out is tricky. Google's goal is to make almost everything serverless and capable of autoscaling -- abstracting cluster management and configuration away from the customer, Hopkins said.
In addition, Google's recent Apigee acquisition enables API providers to design, secure, deploy, monitor and scale APIs. Google also provides Kubernetes to automate deployment, scaling and management of containerized applications -- a capability "no one else can touch," Hopkins said.
Still, AWS data management services hold the lead in terms of compatibility with existing standards and platforms. While Amazon Aurora is compatible with MySQL and many existing applications, Google's Cloud Spanner, a "very interesting new database that combines scalable writes and reads with ACID [atomicity, consistency, isolation and durability] transactions," requires application rewrites, Binsztok said.
Hopkins agrees. "The strength of AWS is in its pace of innovation," he said. "They are the leader in rolling out new capabilities, and Microsoft is not innovating as quickly."
AWS' weaknesses are data unification, security, metadata and governance, but Hopkins said the cloud provider is addressing these issues. The cloud provider needs to be clear on how to apply policies and data security rules across all tools, streaming and batch, in real time, Hopkins said. "Right now, there really isn't any way to do that."
In contrast, Microsoft's Azure caters to the enterprise. "[Azure] is still fairly immature, but they have more of these capabilities than AWS," Hopkins said. "They have a common data catalog, and they are looking to beef it up with metadata management and security governance." Microsoft also leads in usability with its Azure Data Lake service, he added.
Keep costs in mind
Cloud providers also compete in data storage pricing, and the market doesn't benefit Microsoft.
"When I talk to customers about Microsoft, they say [the cloud provider's] pricing structure is a little out of whack -- on the high side," Hopkins said. While these enterprises want to go with Microsoft, it's difficult to do in an affordable way. And Microsoft won't ever budge on pricing, so they don't choose Azure, he said.
AWS data management services, on the other hand, offer varied pricing options, such as Spot instances and other features, to help reduce unnecessary costs. "They aren't extremely flexible either, but they tend to be more favorable to customers," Hopkins said.
Google's pricing structure is particularly murky. "They are used to working in a Silicon Valley mindset, so there are issues in billing," Hopkins said. "Their billing is a mess, and clients tell me they have to get a whole lot better."
Understand data analytics options
When it comes to data management, AWS has fewer gaps than its competitors, according to 451 Research analyst Carl Brooks. Each of the cloud providers has a fairly well-developed portfolio that spans between raw data storage and database-as-a-service offerings.
"AWS has some business intelligence services where you can provide a lot of traditional data, such as Excel spreadsheets, and generate all kinds of charts and graphs," Brooks said.
For more sophisticated data analytics needs, Google is the most exciting place to play, Brooks noted. "It's the kind of environment where advertisers can combine all kinds of data and apply tools like Hadoop and MapReduce to accomplish what they need," he said. Azure also has options to manipulate unstructured data, but its portfolio generally lags behind AWS.
Many customers choose familiarity and compatibility over analytics capabilities or a provider's individual breadth of services, Brooks added. "Folks more accustomed to Microsoft SQL will be drawn to Azure SQL, and those more familiar with MySQL and other open source options will probably gravitate to AWS," he said.
All you need to know about AWS data management options
AWS data management gap can be closed
Here are some non-AWS options for data management