Hi,
I’m trying to find out the answer for this: How memory row tables are handled in case of redundancy_level = 2?
Are the redundancy partitions get stored in memory also?
So if i have a cluster of 2 nodes, 16GB each, in redundancy_level=1 i’ll have 32GB of “storage” for memory tables, in case of redundancy_level =2 will this stay 32GB or will it halve to 16GB?
In redundancy_level = 2, in-memory rowstore tables are stored in memory on both master and slave partitions. We will likely support an option to only store them on disk on slave partitions at some point in the future (at the cost of slower failovers).
Yep. For larger tables you can try using columnstore. Our columnstore does a better job then other columnstores at selective filters and concurrent writes (something traditionally rowstores are much better at).