Score:0

What is the recommended configuration for lower latency(TTFB) for Route 53 for CloudFront in future proof way

my flag

I'm new to AWS.

I use CloundFront for host a fully static website under custom domain name + https. Im trying to understand [without success] how to configure Route 53 to get best result location-wise for a single country and make it work for others.

The one page of documentation tells

You must create an Alias record for the CloudFront distribution to work.

which is pretty understandable on the surface. However it does not recommend routing policy for setting up aliased record.

According to another piece of documentation latency based routing seems like the best option given CloudFront edges are indeed in multiple locations:

Latency routing policy – Use when you have resources in multiple AWS Regions and you want to route traffic to the region that provides the best latency. You can use latency routing to create records in a private hosted zone.

But when it comes to practice the AWS console allows to configure to single latency record per region: enter image description here and there is no option to configure them all at once and even if i want to configure for a single country the third piece of documentation describing best practices tells the following:

When using geolocation, geoproximity, or latency-based routing, always set a default, unless you want some clients to receive no answer responses.

According to this description if I going to configure an alias latency record to lets say Germany it would return blanks for every not configured region, and even if I configure it for all regions AWS may decide to offer new location and it serve nothing because there is no way to set default for latency alias record.

Question: What are the best practice to configure record in Route 53 for CloudFront in future proof way given the info from the documentation?

Score:1
mw flag

You just have to create the Route53 alias to your CloudFront distribution, nothing more - as the CloudFront network should be used for requests from any region. It will automatically route to the best (mostly nearest) edge location within the configured price class of your distribution (or sometimes outside it, in that case you pay the lowest fees available in the selected class).

The latency routing policies are for custom setups, where you self-manage clusters in different regions.

Igor B avatar
my flag
Interesting! Clusters of what (given the CloudFront is global)?
Marces Engel avatar
mw flag
One example would be ECS clusters for container-based servers which you'd like to distribute around the globe, say you want to have a one cluster for US based customers (in one of the US regions) and one for EU based customers in one of the EU regions.
Marces Engel avatar
mw flag
If this post answered your question, feel free to accept it as the answer. Much appreciated!
mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.