Our SaaS product deployment uses a two mirror(serving / non serving) architecture where the two mirrors don't share anything (database / app services are all isolated between mirror in two different AZs) within each mirror is deployed with infrastructure/services in each AZ。 This includes ec2 instances(mirror 1 would have ec2s from AWS AZ1 vs mirror two will have in AZ2 as an eg). We divert traffic automatically based on health assessment of the mirror and failover between mirrors. I am leading an effort to add ElastiCache / Redis into the system. I have two options:
- Uses HA clustered more where mirrors use a single HA cluster
- Two separate mirror specific elastic clusters.
Any opinions on folks who are experienced with running in production ElastiCache in large scale on cost/ benefit on sharing / not sharing.?
Our SaaS product deployment uses a two mirror(serving / non serving) architecture where the two mirrors don't share anything (database / app services are all isolated between mirror in two different AZs) within each mirror is deployed with infrastructure/services in each AZ. This includes ec2 instances(mirror 1 would have ec2s from AWS AZ1 vs mirror two will have in AZ2 as an e.g.). We divert traffic automatically based on health assessment of the mirror and failover between mirrors.
I am leading an effort to add ElastiCache / Redis into the system. I have two options:
- Uses HA clustered more where mirrors use a single HA cluster
- Two separate mirror specific elastic clusters.
Any opinions on folks who are experienced with running in production ElastiCache in large scale on cost/ benefit on sharing / not sharing.?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论