There’s no one-size-fits-all answer unfortunately. How much data do you have? Do you use a lot of SQL? A lot of compute? What are your resilience requirements?
For an “average” deployment I’d start looking at the “Memory optimised” instances (r5 and r5a). Of course, no one has an average deployment. As ever, a good place to start is with the capacity planning guide in the documentation: https://apacheignite.readme.io/docs/capacity-planning Regards, Stephen > On 21 Jun 2019, at 09:56, brendfox <brendan....@per.gov.ie> wrote: > > Hi All, > > Does anyone have recent experience of deploying Apache Ignite on AWS EC2 > instance cluster? > There is much variance in the AWS EC2 instance types in terms of number of > CPUs (therefore performance), memory availability and network bandwidth. > Ultimately costs would increase with the higher spec AWS instance types. > > Would anyone have a recommendation on the minimum resource requirements to > stand up a production Ignite cluster on AWS? That is what is an ideal > resource specification from recent production experience. Some alternative > solutions have provided this information in advance with the AWS Instance > type recommended being more towards the premium end for CPU, memory and > network bandwidth. > > Thanks , Brendan > > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/
