Interesting blogs

[Nitin]

Hi !

Some interesting blogs in design and performance of db:

http://highscalability.com/blog/2012/8/14/memsql-architecture-the-fast-mvcc-inmem-lockfree-codegen-and.html

http://dom.as/2012/06/26/memsql-rage/

Regards,

Nitin

[Rami Ojares]

Another marketing propaganda debunked.

- rami

[Nitin]

I am not linked to both in anyways :) Thought it would be useful read for developers if you read including comments there.

[Rami Ojares]

> Thought it would be useful read for developers if you
read including comments there.

What exactly did you find in those discussions useful?

- rami

[Nitin]

The reasoning for their design and data structures.

[Thomas Mueller]

Hi,

Thanks, it is quite interesting. MemSQL seems to use skip lists instead of the regular B-trees. When I wrote HypersonicSQL, I actually started with a skip list implementation, that was later replaced with an AVL tree and now a B-tree. MemSQL compiles the queries using GCC. They use read-committed transaction isolation by default. All quite interesting choices. I'm not sure if all of those are very good choices (only time can tell), but interesting.

Regards,

Thomas

--
You received this message because you are subscribed to the Google Groups "H2 Database" group.

To view this discussion on the web visit https://groups.google.com/d/msg/h2-database/-/GOfyAUnYQjcJ.

To post to this group, send email to h2-da...@googlegroups.com.
To unsubscribe from this group, send email to h2-database...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/h2-database?hl=en.

[Nitin]

Hi Thomas,

Can you share reasons to change list implementations? In  general was it scaling or performance or code complexity?

Thanks

Nitin

[Thomas Mueller]

Hi,

Can you share reasons to change list implementations?

Do you mean why I switched from the "skip list" to a "AVL tree" and then to a "B-tree"? Yes. I implemented the skip list algorithm because it was new to me (so - not because of concurrency or so). Then I switched to AVL tree because it was faster for in-memory operations. Then I switched to B-tree because it was faster for disk based operations (actually it is not much slower for in-memory operations either). My skip list implementation was relatively slow and used more memory. Some recent research says B-trees are good for in-memory use.

Regards,

Thomas

[Nitin]

Thanks Thomas :)