Re:MySQL is bazillion times faster than MemSQL
by chris_wot on Tuesday June 26, 2012 @10:49PM
(#4165268)
Since you never wait on I/O and don't take locks, optimizing execution makes a lot of sense actually. All that's left is execution and network.
I'm not sure where you are getting this information. According to the MemSQL documentation, they have one and only one isolation level, READ COMMITTED. But READ COMMITTED takes write locks and blocks reads, but doesn't take any read locks. Blocking behaviour still (and should!) still occur.
It sounds like you are looking for dirty read behaviour, but from the literature MemSQL doesn't support READ UNCOMMITTED.
Update: Read a later FAQ on the MemSQL website. They are using MVCC, so write locks are irrelevant. My apologies to the MemSQL team.
Re:MySQL is bazillion times faster than MemSQL
by vasilizo on Tuesday June 26, 2012 @11:25PM
(#4165345)
Chris, MemSQL utilizes versioning to implement READ COMMITTED. In this implementation readers are never blocked.
It is addressed in this FAQ: http://developers.memsql.com/docs/1b/faq.html#c1-q4.
Re:MySQL is bazillion times faster than MemSQL
by chris_wot on Wednesday June 27, 2012 @12:36AM
(#4165539)
OK, didn't see that you are using multiversion concurrency control.
I didn't realise you had a new FAQ up on the website - I'll have a look.
Re:MySQL is bazillion times faster than MemSQL
by tintor on Tuesday June 26, 2012 @11:51PM
(#4165412)
@Criss
Isolation levels are used to specify semantics only. DBMS is free to implement them in different ways.
Take a look at this VLDB 2012 paper about optimistic concurrent control in main memory databases, like MemSQL, which also describes implementation of isolation levels:
http://arxiv.org/pdf/1201.0228v1.pdf
There are use cases where 50ms is way too long, and where jitter associated with disk I/O is not acceptable. Capital markets for example. You are looking for sub-millisecond latency to be in the intraday post-trade game, and single digit milliseconds for quote-to-trade. The number of instructions in the hot code path (and cycles per instruction) actually starts to matter. Lock-free sturcutres utilized for storage are also important. This combination gives MemSQL some rather unique low-latency properties that certain folks can exploit to their delight.
To your other point, random reads and key-value lookups is where MemSQL shines. Since you never wait on I/O and don't take locks, optimizing execution makes a lot of sense actually. All that's left is execution and network.