David Ribeiro Alves has posted comments on this change. Change subject: Kudu Consistency Blog Post Pt1 ......................................................................
Patch Set 10: (37 comments) http://gerrit.cloudera.org:8080/#/c/7019/10/_posts/2017-08-21-kudu-consistency-pt1.md File _posts/2017-08-21-kudu-consistency-pt1.md: PS10, Line 7: what is currently implemented > Consider "current features" instead. Done PS10, Line 7: what are the next step > Consider just "next steps". Done PS10, Line 8: some of > nit: remove Done PS10, Line 8: into > on Done PS10, Line 12: we review/introduce some concepts > This sentence has the same purpose as the last sentence of the previous par I removed the repitition. does it look better now? PS10, Line 23: , > Remove Done PS10, Line 25: single machine > single-machine Done PS10, Line 25: e thread > single-threaded Done PS10, Line 25: > a Done PS10, Line 26: m > add comma Done PS10, Line 26: would happen > happen Done PS10, Line 32: single-threaded, single-machine storage > consider removing? Done PS10, Line 34: s > add comma Done PS10, Line 35: , which > This sentence is getting long; consider starting a new one. Done PS10, Line 36: in > on Done PS10, Line 37: having to deal > that have to deal Done PS10, Line 39: ' > Is this a well-known quote? Done PS10, Line 39: ' > If the idea here is to emphasize, consider using italics or bold, if it's a Done PS10, Line 47: in > unbold the "in" here and below Done PS10, Line 55: , > ; Done PS10, Line 66: that execute > Remove Done PS10, Line 68: , i.e. reads, > Consider surrounding with ()s instead of commas, same below. The commas sor Done PS10, Line 71: with right set > with the right Done PS10, Line 73: (Spanner achieves 5 9's availability) > Not sure if this tidbit is important unless we're planning on bringing it u it serves the purpose of putting the availability sacrifice into context. in spanner's case we still get 5 9's availability Line 73: cases (Spanner achieves 5 9's availability). For the write path, we often made similar choices in Kudu. > of Done PS10, Line 75: informed > directed? Done PS10, Line 79: factor > remove Done PS10, Line 94: much higher in the priority scale > a much higher priority Done PS10, Line 106: future feature > Should we link to a JIRA here? maybe, but apache jira is down for logins. I'll fix in the next iteration if I can add it. PS10, Line 121: kudu > Kudu Done PS10, Line 122: with the particularity that causally related mutations : may be required to have increasing timestamps. > What are "causally-related mutations"? Done PS10, Line 133: more > Remove Done PS10, Line 137: a > remove Done PS10, Line 143: get/set _timestamp tokens_ from/to > coordinate a set of _timestamp tokens_ with clients Done PS10, Line 145: __STRICT SERIALIZABILITY__ > Is there a reason these are caps'ed and bolded? consider lower-case and ita I specifically mentioned that these kinds of concepts would not be introduced in this post, but I'll add a note and references to where the reader can find out more PS10, Line 145: __LINEARIZABILITY__ and : __SERIALIZABILITY__ > perhaps link to a definition? Done PS10, Line 146: g the user pass messages from/to clients : with the timestamp tokens > having the user coordinate the timestamp tokens across clients Done -- To view, visit http://gerrit.cloudera.org:8080/7019 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: Icaec0c8ace20651a65901a8e1786e785265540d1 Gerrit-PatchSet: 10 Gerrit-Project: kudu Gerrit-Branch: gh-pages Gerrit-Owner: David Ribeiro Alves <davidral...@gmail.com> Gerrit-Reviewer: Alexey Serbin <aser...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: David Ribeiro Alves <davidral...@gmail.com> Gerrit-Reviewer: Will Berkeley <wdberke...@gmail.com> Gerrit-HasComments: Yes