Using SaaS and Cloud Require Careful Data Wrangling
The annual AWS re:invent conference wrapped up this week, with more than 43,000 attendees gathering. Some hot topics included AI and machine learning, but the bigger picture revolved around Amazon’s work to continue its public cloud dominance, according to this account. Microsoft and Google continue to work to take market share away, though AWS had a lot of announcements to make. Amazon announced new releases in the instances category, plus more choice in running databases in the cloud as well as IoT device management capabilities. Other releases and developments of interest are private network connection availability with inter-region VPC peering and a continuous threat detection service called GuardDuty.
Also from AWS are some tips on preventing cloud or SaaS app data loss. Adopting any SaaS application means you’re moving and sharing the data involved with that app. Some of these tips on avoiding data loss overlap closely with one of our favorite tips: Negotiate, read and demand enforcement of your SLA with any and all providers. That SLA dictates data backup and data security, and you shouldn’t assume that a provider is backing up or securing your data in the way that’s appropriate for your business. Data recovery is also an important question for any business using cloud or SaaS. It’s also very possible that providers don’t mention these major topics at all in an SLA, which brings us back to the common advice we give: Know what to expect from your provider with any SaaS or cloud application, and keep pushing for more performance-based SLAs.
On the topic of cloud-heavy, forward-looking computing infrastructures, serverless computing continues to find curious users. Containers made a splash a few years ago with their ability to abstract software in between the OS and application code, and are gaining market momentum around the Kubernetes platform. Serverless computing, though, is also making strides. Serverless, or FaaS, are useful for applications that don’t have continuous needs, and could help IT teams abstract infrastructure management entirely. And of course, the two can be used together. As we’ve seen, on-premises infrastructure isn’t going away, even as cloud computing is adopted more and more. IT teams will have plenty of options for those clouds and software-defined tools.
Getting data to various cloud deployments is a challenge for lots of IT teams. One recent survey found that the biggest challenges to cloud adoption are updating apps to connect to the cloud, and integrating object storage (like Amazon’s) into their infrastructure. The Amazon S3 protocol isn’t a natural fit for enterprise applications, so this presents a major challenge for IT teams who are both using AWS for public cloud and maintaining on-premises data centers—likely a majority of companies. Solving these challenges will likely require a hard look at the data a company is collecting, and decisions around its importance and future use, such as how often archival data needs to be accessed.