On Wed, Jun 8, 2011 at 11:24 PM, Robert Haas <robertmhaas@gmail.com> wrote:
> I did a brief literature search for papers on breaking the
> WAL-serialization bottleneck today and hit upon this:
>
> Aether: A Scalable Approach to Logging, Ryan Johnson, Ippokratis Pandis, et al.
> http://infoscience.epfl.ch/record/149436/files/vldb10aether.pdf
>
> Section 5 appears to be the most relevant to our problems with WALInsertLock.
>
> They reject the approach that I proposed, wherein WAL is generated by
> individual backends in their own queues and serialized later: "While a
> total ordering is not technically required for correctness, valid
> partial orders tend to be too complex and interdependent to be worth
> pursuing as a performance optimization"; see also appendix A.5, which
> may be succinctly summarized as "no one does that".
heh -- makes total sense. great stuff.
merlin