Question on sharding

Prior to IRIS, using ECP to share databases under heavy I/O load has known latency issues (in our environment), which pretty much restricted using shared database with relatively static or slow-changing data. In IRIS, will sharding mitigate the latency issue and allow any table to be shared?

David

Dmitry Maslennikov · Nov 17, 2019

Sharding will allow you to split your data between physically different servers with each own storage. So, you may get faster simultaneous read/write of data. At the same time with ECP configuration, your application servers will get direct access to the sharding node which holds data.

So, if it is correctly configured you should get some profit from it, for sure.

1 0

Benjamin De Boe · Jul 15, 2020

Hi David, maybe you can elaborate a little more on the particular latency challenge you faced?

As Dmitriy mentioned, sharding will spread your data and corresponding query workload across multiple nodes in order to achieve higher efficiencies on very large datasets and is especially fit for read-mostly workloads. ECP, when used in a typical application server setup, is meant to distribute user-bound workload across multiple servers, so serving a slightly different goal. Depending on your use case, either of those (or a combination of them) can be more appropriate. See also this overview and introductory video for more info.

0 0