Record limits - Even for simple aggregations!

When executing a simple “select count(*) from table”, the return payload simply states “10000”. The underlying source is Apache Hive. The same behaviour exists for a virtual table on top of a CSV file in HDFS. What gives??

@greendisc

Not able to follow exactly. Are you saying you get different results via Dremio?

Preview queries are run one a subset of data for UI queries. You need to switch to RUN to get a full result set:

image

Yes! That’s the issue. I was using the keyboard shortcut “ctrl-enter”, which in the web SQL world is universal (almost) for executing queries. Not used to “ctrl-shift-enter”! Is this a setting that can be modified?

Since runs can be expensive, we default to previews. There isn’t currently a way to switch the default behavior and will open a ticket to consider it. We are working on improving the preview/run behavior to be more obvious.

Gotcha, thanks for the quick responses.