Hello Antony
Thank you for your detail response that matches more or less my understanding. We will pay attention to the bucket count in our tests Thank you one again Best regards Paul From: Anthony Baker <[email protected]> Sent: 15 October 2018 15:17 To: [email protected]; [email protected] Subject: Re: Increase Region size dynamically To manage the growth of data, you have a couple of options: 1) Consider the data lifecycle and what data needs to be kept live in-memory. Perhaps you could use eviction policies with overflow to disk and/or CacheLoaders to balance performance and memory use. 2) For replicated regions, storing larger amounts of data means larger heap sizes on each member. You would need to stop each Geode process and restart with a different heap size. 3) For partitioned regions, you can scale horizontally by just adding more members to the cluster and rebalancing. You’ll want to ensure your bucket count (default 113) is much larger than your cluster size. Anthony On Oct 14, 2018, at 7:17 AM, Paul Perez <[email protected]> wrote: Hello All, In one of our project, we are unable to forecast the size of the distributed regions in the next months. So, after a few months on production, the region size could be increased on multiply by 2-3-N times. To solve this issue, * Do we need to duplicate the increasing region to reduce the size of each region instances and consequently increase the admin tasks? * Do we have to stop the server, resize the server? * Or is there another way to do it dynamically and increase the memory size Thank for your help paul
