13 - Query Execution II (CMU Databases Systems / Fall 2019)

[@kingofgaming9164]

I hope they did well on the exam.

[@jianhe]

in 36:55, I am confused. Select * from A where A.value > 99; You first step1 is bring table A from disk to memory then do the filter operation. I don't think parallelism here.

[@shuheikitagawa8472]

Finally, I can hear hit it again.

[@blacknick3931]

We can have Coroutine per DBMS Worker I guess.

[@franciscomuniz4475]

For distributed systems, the query planning is usually the same? the gather, repartition and distribute looks like map and shuffle from spark

[@andypavlo]

Because it is the same. This terminology comes from the early distributed databases from the 1980s + 1990s, well *before* Spark was invented. See these papers: dl.acm.org/citation.cfm?id=1559865 people.eecs.berkeley.edu/~brewer/cs262/5-dewittgray92.pdf

[@longwang2733]

'coroutine' tech should be mentioned in this part?

[@ManikumarAryas]

Teradata is massively parallel processing database. Any specific reason for not mentioning Teradata anywhere in your classes? I dint see even a single instance so far as such.

[@AshishNegi1618]

lol.. he mentions it in next lecture.. around 3:20 kzbin.info/www/bejne/aGnLhImhhNhprpo

[@luojihencha]

Why that guy has so many questions????

[@HarshKapadia]

Why not!?

[@eytehs]

because he has a fully functioning brain

[@luojihencha]

Some of his questions are really stupid. His classmates all have well functioning brain and they don't ask so many stupid questions every single class.

[@eadrenking3960]

就你个中国人事儿多，偷学老外课程，话还贼多