Artwork

コンテンツは Lee Atchison によって提供されます。エピソード、グラフィック、ポッドキャストの説明を含むすべてのポッドキャスト コンテンツは、Lee Atchison またはそのポッドキャスト プラットフォーム パートナーによって直接アップロードされ、提供されます。誰かがあなたの著作物をあなたの許可なく使用していると思われる場合は、ここで概説されているプロセスに従うことができますhttps://ja.player.fm/legal
Player FM -ポッドキャストアプリ
Player FMアプリでオフラインにしPlayer FMう!

News Report: EKS Pricing

3:15
 
シェア
 

Manage episode 294678598 series 2937999
コンテンツは Lee Atchison によって提供されます。エピソード、グラフィック、ポッドキャストの説明を含むすべてのポッドキャスト コンテンツは、Lee Atchison またはそのポッドキャスト プラットフォーム パートナーによって直接アップロードされ、提供されます。誰かがあなたの著作物をあなたの許可なく使用していると思われる場合は、ここで概説されているプロセスに従うことができますhttps://ja.player.fm/legal

Amazon recently announced a reduction in the cost of their Elastic Kubernetes Service. They lowered the cost of the Amazon EKS service by 50%. The new pricing is greatly appreciated, but is still a barrier to entry in using the EKS service for small clusters.

Links and More Information

The following are links mentioned in this episode, and links to related information:


Special Report

Amazon recently announced a reduction in the cost of their Elastic Kubernetes Service.

As of January 21, 2020, the price of Amazon EKS will be lowered by 50%. The previous price for using the EKS service was 20 cents per hour for each Kubernetes cluster managed. This works out to approximately $144 per month. The new price for running an EKS cluster will now be 10 cents per hour, which works out to approximately $72 per month.

It should be noted that this is the price necessary to operate the cluster, regardless of the number of workers contained in the cluster. You must still pay for the computation costs — either EC2 instance hours or Fargate compute resources — in order for the cluster to be useful. The cost reduction is for the “overhead” cost of maintaining the cluster, which covers the Amazon cost of managing the cluster.

This price has always been an issue with the EKS service. The non-Kubernetes Amazon container management service, ECS, does not have this overhead cost — all you pay for is the cost of running the workers. Only EKS has this overhead. So, the reduction of this overhead cost should be seen as significant reduction in the barrier to entry in using the EKS service.

While I applaud AWS for taking this step, I still believe this overhead cost is too high. It still makes EKS less attractive than ECS for small clusters. The cost is not an issue for large clusters, since you can amortize the cost out over the size of the cluster. But if you run a large number of small clusters, this overhead cost can be significant. For example, if you are running 100 clusters, the overhead cost is around $87,000 per year. Overhead. Now that’s a lot less than the $175,000 it was before the price change, but it’s still quite a bit of overhead.

That said, I applaud AWS for making this price change. It’s a consistent strategy they use — pass on cost savings to their customers as their costs go down due to volume. It’s a strategy very much appreciated by their customers.

A link to the announcement is given in the show notes, along with a link to the EKS and ECS services.


This podcast uses the following third-party services for analysis:
Chartable - https://chartable.com/privacy
Podtrac - https://analytics.podtrac.com/privacy-policy-gdrp
  continue reading

31 つのエピソード

Artwork
iconシェア
 
Manage episode 294678598 series 2937999
コンテンツは Lee Atchison によって提供されます。エピソード、グラフィック、ポッドキャストの説明を含むすべてのポッドキャスト コンテンツは、Lee Atchison またはそのポッドキャスト プラットフォーム パートナーによって直接アップロードされ、提供されます。誰かがあなたの著作物をあなたの許可なく使用していると思われる場合は、ここで概説されているプロセスに従うことができますhttps://ja.player.fm/legal

Amazon recently announced a reduction in the cost of their Elastic Kubernetes Service. They lowered the cost of the Amazon EKS service by 50%. The new pricing is greatly appreciated, but is still a barrier to entry in using the EKS service for small clusters.

Links and More Information

The following are links mentioned in this episode, and links to related information:


Special Report

Amazon recently announced a reduction in the cost of their Elastic Kubernetes Service.

As of January 21, 2020, the price of Amazon EKS will be lowered by 50%. The previous price for using the EKS service was 20 cents per hour for each Kubernetes cluster managed. This works out to approximately $144 per month. The new price for running an EKS cluster will now be 10 cents per hour, which works out to approximately $72 per month.

It should be noted that this is the price necessary to operate the cluster, regardless of the number of workers contained in the cluster. You must still pay for the computation costs — either EC2 instance hours or Fargate compute resources — in order for the cluster to be useful. The cost reduction is for the “overhead” cost of maintaining the cluster, which covers the Amazon cost of managing the cluster.

This price has always been an issue with the EKS service. The non-Kubernetes Amazon container management service, ECS, does not have this overhead cost — all you pay for is the cost of running the workers. Only EKS has this overhead. So, the reduction of this overhead cost should be seen as significant reduction in the barrier to entry in using the EKS service.

While I applaud AWS for taking this step, I still believe this overhead cost is too high. It still makes EKS less attractive than ECS for small clusters. The cost is not an issue for large clusters, since you can amortize the cost out over the size of the cluster. But if you run a large number of small clusters, this overhead cost can be significant. For example, if you are running 100 clusters, the overhead cost is around $87,000 per year. Overhead. Now that’s a lot less than the $175,000 it was before the price change, but it’s still quite a bit of overhead.

That said, I applaud AWS for making this price change. It’s a consistent strategy they use — pass on cost savings to their customers as their costs go down due to volume. It’s a strategy very much appreciated by their customers.

A link to the announcement is given in the show notes, along with a link to the EKS and ECS services.


This podcast uses the following third-party services for analysis:
Chartable - https://chartable.com/privacy
Podtrac - https://analytics.podtrac.com/privacy-policy-gdrp
  continue reading

31 つのエピソード

すべてのエピソード

×
 
Loading …

プレーヤーFMへようこそ!

Player FMは今からすぐに楽しめるために高品質のポッドキャストをウェブでスキャンしています。 これは最高のポッドキャストアプリで、Android、iPhone、そしてWebで動作します。 全ての端末で購読を同期するためにサインアップしてください。

 

クイックリファレンスガイド