4 ms·

Hmm, tell us more? Databases are generally known for their good performance (under locking conditions - this is the relevant factor I guess?), after all...

• redis_mlc 7y ago
  > Databases are generally known for their good performance If the data is on a filesystem, then sed, grep and cut pipelines will likely be your fastest option (Yahoo! processed petabytes of logfiles for decades that way.) If the data is already inside a database table and indexed well, that could be fast enough. But generally speaking, the ETL is often a bottleneck. And DBAs are $$$$ compared to "the UNIX way." Source: DBA