3 ms·
Expensive but very fast tape with easy random rewind, but still largely sequential write access, is something that spark / hadoop could readily use.
by nine_k 7y ago
Expensive but very fast tape with easy random rewind, but still largely sequential write access, is something that spark / hadoop could readily use.