Hi @sumit -
We don’t keep highly detailed table-level statistics at the present,
as our usage model is different than a transactional database. For example, when using Immerse with a dataset, a simple crossfilter might trigger 5-10 queries as filters are applied. In that case, queries executed per hour doesn’t really capture the usage pattern. Same with the CPU/GPU intensity, as we default to using all available resources to maximize throughput for each query, as opposed to minimizing resource usage for multi-tenancy.
\memory_summary will get you some of the information you are looking for,
but for the most part, these types of statistics aren’t applicable to the current MapD architecture.
After discussing internally, we may start to provide more of this type of information as internal tables, so that people can better understand their resource usage. But using
\o in mapdql to optimize your table DDL and
memory_summary are the two main approaches right now for optimizing resource usage.