ZhangYao has posted comments on this change. ( http://gerrit.cloudera.org:8080/13412 )
Change subject: KUDU-2826: Add tail for mutation list to accelerate update in memrowset ...................................................................... Patch Set 3: (1 comment) A bout the user time exceed read time, I try to feature out why but still don't know. I find out that kudu use RUSAGE_THREAD to just count the current thread time and this is supposed to lead to read time > user time, so I have no idea. Following is my experiment environment may be help to figure out this question: Host: MacOX Virtual Machine: Docker with centos7(core 4.9.125-linuxkit) I run kudu in docker on Mac OX. What's more the NTP configuration was not right in docker, does that matter? By the way, I am doing the ycsb benchmark, after that will update the result and code. http://gerrit.cloudera.org:8080/#/c/13412/3/src/kudu/tablet/mutation.cc File src/kudu/tablet/mutation.cc: http://gerrit.cloudera.org:8080/#/c/13412/3/src/kudu/tablet/mutation.cc@54 PS3, Line 54: void Mutation::AppendToListAtomic(Mutation **redo_head, Mutation **redo_tail) { > This is no longer "atomic" in that redo_head and redo_tail may disagree (if Agree with you. Because kudu writes the row with lock, so this function will not be enter by more than one thread for the same row. Redo_tail only accessed by writer and redo_head only accessed by reader.For those I think the "disagreements" will not matters. -- To view, visit http://gerrit.cloudera.org:8080/13412 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I945f332f8241ecb3cc6ab66e67ee3ee2b4e49be8 Gerrit-Change-Number: 13412 Gerrit-PatchSet: 3 Gerrit-Owner: ZhangYao <[email protected]> Gerrit-Reviewer: Adar Dembo <[email protected]> Gerrit-Reviewer: Andrew Wong <[email protected]> Gerrit-Reviewer: Grant Henke <[email protected]> Gerrit-Reviewer: Kudu Jenkins (120) Gerrit-Reviewer: Todd Lipcon <[email protected]> Gerrit-Reviewer: Yao Xu <[email protected]> Gerrit-Reviewer: ZhangYao <[email protected]> Gerrit-Comment-Date: Tue, 28 May 2019 10:19:54 +0000 Gerrit-HasComments: Yes
