Route 53: AWS’ Secret DNS Cloud Sauce

Route 53: AWS’ Secret DNS Cloud Sauce

The most frequently-discussed elements of Amazon Web Services (AWS) are its EC2 virtual server options and S3 cloud storage. Yet some long-term AWS customers reserve their highest praise for a rather more obscure product: the Route 53 scalable DNS management service.

Road picture from Shutterstock

AWS ANZ managing director Ed Lenta describes the role of AWS for many organisations as “undifferentiated heavy lifting”. That’s a description that works well for Route 53, which allows businesses to perform complex traffic redirection and management across AWS-based cloud projects.

Matt Barrie, CEO of, is a particular fan. Speaking at an AWS media lunch last week, he noted that Route 53 “doesn’t get attention” compared to other cloud offerings. However, for, it’s “absolutely mission critical”, he said.

The main benefit of Route 53 for is the ability to manage traffic to its various country domains. It operates numerous local sites (such as or, but these all draw on the same central pool of data. “We have country domains around the world,” Barrie said. “Without Route 53, it would be very awkward.”

Another common use case for Route 53 is to redirect traffic automatically to the most relevant Amazon region, Lenta noted. Amazon operates eight distinct data centre regions (not include a government-only option offered in the US), including one based in Sydney that was launched in November last year. While the cheapest rates usually apply to its US data centres, many customers prefer to use other options for latency, compliance or traffic cost reasons.

The pricing model for Route 53 is based on the number of queries made to the service. Customers pay $US0.50 per “hosted zone” per month, and then an additional $US0.50 per million queries made to the service. Queries which are routed based on latency to different regions are slightly costlier, at $US0.75 per million queries.

Unlike many AWS options, the basic service charges aren’t pro-rated across the month, but instances set up for testing purposes and deleted within 12 hours aren’t charged.


  • Yep Route 53 is boss. The only down side is you have to use their custom name servers, which are selected on create of your dns entries, so you can never know the nameservers ahead of time. A bit painful if you want to use Route 53 for client website hosting, or if you want to change to a different DNS provider in the future.

  • This will come in handy to combat their terrible downtime and unreachable servers problems……..

    .. I honestly continually wonder why anyone even cares about AWS – it seems to be mostly because they have good marketing..

    • I’m surprised with your comment as we’ve had fantastic success and achieved very high availability with AWS. Amazon provide some excellent guidance on how to achieve high availability across zones and regions and it’s really up to the implementer to ensure they put in the appropriate redundancy for their requirements. Even with this said, we’ve had very little reason to rely on this redundancy over the last two years as their availability has been very good.

    • I think most of the bad experiences with AWS come from people expecting it to be like a standard VPS provider. It’s not, and you need to do your homework before deciding if it’s suited to your needs.

      If you want reliability, go large/extra large instances and provisioned IOPS, then from what I’ve seen your issues will be minimal.

      • @voges too; no not at all. I think a lot of it stems from how they actually measure uptime – which is not started at all until an entire region is unavailable for at least 5 minutes..

        It might not matter at all if you’re hosting – but when you’re trying to run real time critical applications where downtime costs money, thus why they are hosted in the cloud in the first place..

Show more comments

Comments are closed.

Log in to comment on this story!