Netflix makes use of Ribbon, which is in their terms a "client-side load-balancer". What are the use-cases and advantages of a client-side load-balancer compared to a traditional load-balancer? Is Ribbon, and other Netflix OSS services AWS -specific or can they be used in other contexts?
相关问题
- How to generate 12 digit unique number in redshift
- Use awslogs with kubernetes 'natively'
- Assume/switch role in aws toolkit for eclipse 2.0
- 'no SavedModel bundles found!' on tensorfl
- Installing Python dependencies in AWS Codestar wit
相关文章
- Right way to deploy Rails + Puma + Postgres app to
- how many objects are returned by aws s3api list-ob
- AWS S3 in rails - how to set the s3_signature_vers
- Passthrough input to output in AWS Step Functions
- I cannot locate production log files on Elastic Be
- ImportError: cannot import name 'joblib' f
- Static IP for Auto Scale in AWS
- Step function exceeding the maximum number of char
As already mentioned by using Ribbon you don't need an additional load balancer. You can also configure the balancing algorithm for each client differently, if you need to. At first I was a bit surprised also, but it does make a lot of sense.
Currently Spring Cloud, which uses Netflix OSS for most part, is independent of AWS.
I have made a small POC that you can check out here if want.
One reason for using client-side load balancer can be performance. With client side balancer you can directly contact desired service with one network hop (after initial discovery of course); with traditional load-balancer you need two hops - see my very unprofessional test.
Unfortunately I don't have that much experience with Netflix OOS, I just deduce from documentation that at least Spring Cloud Netflix can run also e.g. on Cloud Foundry.