Mike Percy has posted comments on this change. ( http://gerrit.cloudera.org:8080/8842 )
Change subject: [docs] Add scaling guide ...................................................................... Patch Set 1: (17 comments) This is a great and useful doc. I did a first pass, and while I didn't validate the numbers and have various "usability" related comments. http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc File docs/scaling.adoc: http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@1 PS1, Line 1: [[scaling]] name this file scaling_guide.adoc for SEO reasons http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@2 PS1, Line 2: Scaling Scaling Guide http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@13 PS1, Line 13: Add a blurb at the top of this document describing the purpose of the document http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@46 PS1, Line 46: / nit: missing space before slash http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@57 PS1, Line 57: Memory Memory Consumption? http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@61 PS1, Line 61: baseline rough estimate http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@62 PS1, Line 62: limit approximate memory requirements http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@66 PS1, Line 66: Scaling Is scaling the right word here? http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@68 PS1, Line 68: Data on disk "Memory required per TB of data on disk" ? http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@68 PS1, Line 68: / Here and below, i mentally read this as "divided by" when you mean "per". I think we should write out "per" here. http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@95 PS1, Line 95: Verifying a Memory Limit Verifying if the memory limit is sufficient? http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@129 PS1, Line 129: | File cache | 40% of process maximum | Percentage of maximum allowed open fds reserved for use by can we indicate how this can be tuned? http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@133 PS1, Line 133: / per http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@150 PS1, Line 150: / cold replica how about: * 1600 cold replicas http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@151 PS1, Line 151: | 200 hot replicas | 2 / segment * 10 segments/hot replica * 200 hot replicas + 1 / index * 10 indices / hot replica * 200 hot replicas = 6000 fds having a hard time parsing this expression; consider writing it like (2 fds * 10 segments * 200 hot replicas) + (1 fd * 10 indices * 200 hot replicas) = 6000 fds except frankly i'm not sure how you're supposed to know how many WAL indexes there are and you might need to state some assumptions here given that we say there are 200 hot replicas http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@177 PS1, Line 177: / nit: per, here and below http://gerrit.cloudera.org:8080/#/c/8842/1/docs/scaling.adoc@183 PS1, Line 183: / per -- To view, visit http://gerrit.cloudera.org:8080/8842 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I38d8999addc41fe0b726342a27dbba199ddf7dd2 Gerrit-Change-Number: 8842 Gerrit-PatchSet: 1 Gerrit-Owner: Will Berkeley <[email protected]> Gerrit-Reviewer: Jean-Daniel Cryans <[email protected]> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <[email protected]> Gerrit-Reviewer: Will Berkeley <[email protected]> Gerrit-Comment-Date: Fri, 15 Dec 2017 23:25:32 +0000 Gerrit-HasComments: Yes
