Practise the standard verbs for federating queries across data sources in Presto.
0 / 5 completed
1 / 5
Fill in: 'We ___ a new connector to Presto's catalog so a single query can join data living in an object store with data living in a relational database.'
We 'add a connector' — the standard, simple collocation for wiring a new data source into a federated query engine. The other options are less idiomatic here.
2 / 5
Fill in: 'Federating a query across a slow legacy source without pushing filters down first can ___ the entire cluster waiting on one connector to stream back far more rows than necessary.'
We say a missing pushdown will 'leave' the whole query waiting on one slow source — the standard, natural collocation for the resulting bottleneck. The other options aren't idiomatic here.
3 / 5
Fill in: 'We ___ the query plan before running an expensive federated join, so a filter that could run at the source instead of after the join isn't left unnecessarily late.'
We 'inspect' a query plan — the standard, simple collocation for examining how the engine intends to execute a query. The other options are less idiomatic here.
4 / 5
Fill in: 'We ___ predicate pushdown for a connector explicitly where the underlying source supports it, rather than letting Presto pull an entire table across the network first.'
We 'enable pushdown' — the standard, established collocation for turning on filter execution at the source. The other options aren't the recognised term here.
5 / 5
Fill in: 'We ___ federated query latency against a single-source baseline, so the actual overhead of joining across systems is a measured number, not an assumption.'
We 'benchmark' latency — the standard, established collocation for measuring performance against a reference case. The other options aren't the recognised term here.