Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Triffids
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
2 ms
·
1.
▲
by
Triffids
8y ago
It depends. There are Cloudera kudu engine, it quite easy can handle millions atomic inserts per second. But it can't support true acid transaction. On other hand SQL also can't use true ACID in high load system. Serialazible tran
2.
▲
by
Triffids
8y ago
Generally good idea, analytical queries will be faster than on rdbms. You problem in strange tools and format. Hive = map reduce, JSON = scan for all data. You need to replace map-reduce (Hive) to Impala or Spark 2.x and JSON to columnar fo