Details on the replica update

[Kaloian Manassiev]

Hello,

I recently found out about MySQL Cluster and NDB and I
am impressed.

I would like to know more about the internals and
implementation of NDB. I may suggest that NDB sources
are the best place to start, but just in case someone
knows ... :))

Could someone point me to a technical paper or design
specification where I could find answers to these
particular questions:

 - What is the protocol that NDB uses to propagate
updates to the in-memory tables? When these updates
occur?

 - Also, how is concurrency controlled in an NDB
cluster? In particular, what happens if two different
nodes try to update the same data entry?

 - Is there some sort of distributed locking in
effect, and how track is kept of where locks are in
case of a node failure?

Thank you very much in advance.

Cheers,
Kaloian.


		
__________________________________
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo