Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

I feel like between Redis, S3, Cloud Storage, RocksDB, Cassandra, etc...this area strikes me as one that has been solved as well as we could reasonably expect it to be.

What the world of data needs more of is continued development into novel indexing strategies/implementations. ElasticSearch, Postgres's GIN index on JSONB, MapReduce, graph databases.

I don't need another key value store...

Re: Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

#6

Post by mlevental » Fri, Mar 09, 2018, 1:45 PM UTC

ralusek wrote: post #5

I feel like between Redis, S3, Cloud Storage, RocksDB, Cassandra, etc...this area strikes me as one that has been solved as well as we could reasonably expect it to be. What the world of data needs more of is continued development into novel indexing strategies/implementations. ElasticSearch, Postgres's GIN index on JSONB, MapReduce, graph databases. I don't need another key value store...

there's a new one of those posted like every day - and every single one is "crazy fast" and "super scalable"

Re: Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

#7

Post by dis-sys » Fri, Mar 09, 2018, 2:06 PM UTC

> Totally ordered request processing requires waiting for global consensus at each step, and thus fundamentally limits the throughput of each replica-set.

really? such global consensus increases latency (network round trip plus fsync write), with a fully batched and pipelined concurrent design, when CPU cycles are being saturated in those benchmarks, why throughput is fundamentally limited by such increased latency?

Re: Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

#8

Post by gjem97 » Fri, Mar 09, 2018, 2:18 PM UTC

nickreese wrote: post #3

Looks exciting but very light on details.

There's a lot more detail in the linked paper.

http://db.cs.berkeley.edu/jmh/papers/anna_ieee18.pdf

Re: Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

#9

Post by zellyn » Fri, Mar 09, 2018, 2:18 PM UTC

ralusek wrote: post #5

I feel like between Redis, S3, Cloud Storage, RocksDB, Cassandra, etc...this area strikes me as one that has been solved as well as we could reasonably expect it to be. What the world of data needs more of is continued development into novel indexing strategies/implementations. ElasticSearch, Postgres's GIN index on JSONB, MapReduce, graph databases. I don't need another key value store...

I don't mean this facetiously: did you read the article? This work presents novel ways of "solving" the problem that cause it to beat the other contenders fairly convincingly (in non-linearizable regimes).

Re: Anna: A Fast, Scalable, Flexibly Consistent Key-Value Store

#10

Post by michaelcampbell » Fri, Mar 09, 2018, 2:40 PM UTC

mlevental wrote: post #6

ralusek wrote: post #5

I feel like between Redis, S3, Cloud Storage, RocksDB, Cassandra, etc...this area strikes me as one that has been solved as well as we could reasonably expect it to be. What the world of data needs more of is continued development into novel indexing strategies/implementations. ElasticSearch, Postgres's GIN index on JSONB, MapReduce, graph databases. I don't need another key value store...

there's a new one of those posted like every day - and every single one is "crazy fast" and "super scalable"

Almost as frequent as the new hotness JS framework.