Thank you for answering both my questions. The answers have been of great help!
On Tuesday, April 19, 2016 at 8:34:36 AM UTC-3, Chris Vest wrote: > > Our page cache has nothing to do with the Flash Key-Value Store. The name > similarity is merely a coincidence. > > The LRU behaviour is approximated using a CLOCK-like algorithm, but we > don’t follow any particular published algorithm to the letter. It’s a mix > of inspirations and implementation considerations. > > We don’t mention the algorithm in the documentation because it is an > implementation detail, and we still change and tune it from time to time. > > -- > Chris Vest > System Engineer, Neo Technology > [ skype: mr.chrisvest, twitter: chvest ] > > > On 15 Apr 2016, at 21:53, Telmo Henrique Valverde da Silva < > [email protected] <javascript:>> wrote: > > A few weeks ago I asked on Stack Overflow about a doubt I had concerning > the Neo4j page cache mechanism (which sadly went unanswered here > <http://stackoverflow.com/questions/35789192/when-is-each-implementation-of-the-page-cache-used>), > > but today I have another question about the topic which hopefully is easier > to answer. > > The release notes for 2.2.0 (here > <http://neo4j.com/blog/neo4j-2-2-0-scalability-performance/>) say the > page cache mechanism was implemented based on the LRU-K algorithm, but the > documentation itself doesn't mention said algorithm at all and from what I > gather this implementation has some considerable distinctions from the > original algorithm. > > Considering the *MuninnPageCache* class which implements the *PageCache* > interface on neo4j-io, is this implementation based on the Muninn Flash > Key-Value Store (from this article here > <http://dl.acm.org/citation.cfm?id=2611364>)? If so I don't get where > LRU-K comes into play. > > Thank you for your time. > > -- > You received this message because you are subscribed to the Google Groups > "Neo4j" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected] <javascript:>. > For more options, visit https://groups.google.com/d/optout. > > > -- You received this message because you are subscribed to the Google Groups "Neo4j" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
