2 ms·
Anything you run on your own infrastrucutre I wouldn't really call a BigQuery equivalent. As far as performance goes, if you attacked the problem with the same
by nemothekid 9y ago
Anything you run on your own infrastrucutre I wouldn't really call a BigQuery equivalent. As far as performance goes, if you attacked the problem with the same amount of CPU cores, I think you could get similar performance with Presto/Dremel on top of any fast data store.
What I think is different about BigQuery is just the on-demand parallelism. This means you pay per query instead of per instance which is confusing at first, but then a lot of the operational knobs simply go away (for better or for worse).