A look at news from Amazon Web Services: a backtrack feature and simplified authentication

4 min read

Jul 11, 2018 12:00:00 AM

I recently joined Chris Presley for Episode 5 of his new podcast, Cloudscape , to talk about what’s happening in the world of cloud-related matters. My focus was to share the most recent events surrounding Amazon Web Services (AWS). Some of the highlights of our discussion covered:

Amazon Aurora Backtrack
EC2 C5d Instances
Simplified ALB Authentication
Quicksight Pay-Per-Session Pricing

Amazon Aurora backtrack Amazon Aurora has a new backtrack feature which functions a lot like an “undo” function. When you enable this backtrack feature, it allows you to create a buffer window of up to 72 hours for newly-launched Aurora database clusters . If you make a mistake, say you accidentally drop some tables that you didn’t intend to, you can backtrack the database and it will restore fairly quickly. Keep in mind, however, you cannot enable it for a running cluster - you have to do this when you’re creating the database or when you're restoring. So you’ll need to do a little bit of planning to enable it to backtrack. You will also need to pause your application while you’re doing that. That’s not a surprise, but you will incur some downtime. There is a little bit of a cost involved, and larger windows incur higher costs because you’re storing more change records to disk. But this feature gives you the ability to basically rewind time, because you can backtrack the database to a specific point in time, thus protecting yourself from a bad query. EC2 C5d Iinstances Continuing on the theme of AWS, let’s look at what is new at EC2. In EC2, Amazon launched new C5d instances. This is a big deal because they improved upon the vanilla C5 instances by adding local NVMe storage. The C5ds are designed for computer intensive tasks or workflows such as batch or log processing, distributed / real time analytics, or video encoding. This is for when you need to flush to disk really quickly in order to keep that compute saturated. Really, anything that needs access to a disk and needs to be very fast and low latency. The new C5d instances are going to fill that niche. On these C5ds, any AMI that supports the elastic network adaptor (ENA, NVMe) can be used. So they have support for Amazon Linux, Microsoft 2008 R2, 2012, 2012 R2 and 2016, Ubuntu, RHEL, CentOS and SUSE, which is great. They have a lot of great features, such as encryption, which is baked in. We’re seeing an uptick of people who want to run analytics, and they want that horsepower that keeps things moving. These are available on multiple instance types including spot distances. You can get access to this horsepower, but you don’t need to have them running long term if you don’t need them. This is a fantastic option to fill this void and improves the options for people who want to run these types of workloads. Simplified ALB authentication Application Load Balancers (ALBs) now have integrated authentication logic. This means that you can do your authentication within the ALB and not have to write code to do it in your application. The list of supported technologies is fantastic. It provides integration with Google, Facebook, Amazon (via Amazon Cognito) and OpenID Connect. The gist of this new feature is that the ALB will now check for a session cookie to authenticate users. If users are properly authenticated, their request will be forwarded on with special headers set so the backend application can identify the user and handle authorization. Users who are not properly authenticated receive an HTTP 302 error and are sent on their way. ALB authentication will definitely make life easier, especially for new applications targeting AWS for deployment. This does make Amazon a little bit stickier, but that trade-off is worthwhile for many people who value the convenience of removing yet another aspect of the code base that they’ve traditionally had to manage. This is a huge convenience item for anyone building a mobile app or trying to get something launched very quickly, Quicksight pay-per-session pricing Amazon Quicksight now offers the convenience of billing per session for Enterprise users. This may seem like a small change, but it’s a way to get more people to use Quicksight. To support this, Quicksight users are now broken into three groups: authors, authors and readers. Authors have a monthly cost of $9 a month at the standard tier or $18 a month for the enterprise tier. Admins are super-authors. They can do everything authors can do plus some additional management tasks. Readers can consume dashboards, drill down into data, activate filters and export data in CSV format. The power of these new changes is in the pricing for this tier. Readers pay for access in 30-minute chunks at $0.30 a piece. This price point opens the door to greatly expand the number of people who are consuming Quicksight data without drastically ballooning the monthly price tag. Even better than the per-session pricing is that the total monthly cost per reader is capped at $5, so the maximum amount of spend is predictable. This is a great change and should result in a lot more value for teams that use Quicksight. This was a summary of the Amazon Web Services topics we discussed during the podcast. Chris also welcomed Kartick Sekar (Google Cloud Platform) and Warner Chaves (Microsoft Azure) who also discussed topics related to their expertise. Click here to hear the full conversation and be sure to subscribe to the podcast to be notified when a new episode has been released.

On this page

Ready to unlock value from your data?

With Pythian, you can accomplish your data transformation goals and more.

Speak with Pythian consultants now →

What We Do

A look at news from Amazon Web Services: a backtrack feature and simplified authentication

Share this

Share this

More resources

Deploying Azure Kubernetes cluster with container health monitoring service

Secure the Connection to Your VPC Network from Serverless Environments in Google Cloud

Kubernetes: the easy button for enterprise

Ready to unlock value from your data?