> On Nov. 23, 2015, 7:31 p.m., Yi Pan (Data Infrastructure) wrote: > > docs/learn/documentation/versioned/container/coordinator-stream.md, line 95 > > <https://reviews.apache.org/r/40572/diff/2/?file=1135547#file1135547line95> > > > > Just to make it consistent w/ set-changelog and > > set-container-host-assignment, shouldn't this be: > > {code} > > <td> 'value' => <config-value> </td> > > {code}
My bad. Fixed it. > On Nov. 23, 2015, 7:31 p.m., Yi Pan (Data Infrastructure) wrote: > > docs/learn/documentation/versioned/container/coordinator-stream.md, line 130 > > <https://reviews.apache.org/r/40572/diff/2/?file=1135547#file1135547line130> > > > > Prefer to say "HTTP API" I changed it "an HTTP service". I am not sure if API is the right usage. Let me know if I am mistaken. > On Nov. 23, 2015, 7:31 p.m., Yi Pan (Data Infrastructure) wrote: > > docs/learn/documentation/versioned/container/coordinator-stream.md, line 137 > > <https://reviews.apache.org/r/40572/diff/2/?file=1135547#file1135547line137> > > > > Is this true as for today? I thought that in today's YARN job > > deployment, when AppMaster is down, the whole job is restarted (i.e. the > > containers will be shutdown and restarted as well). The description here is > > more like a "want-to-be" state, not the current state. This is talking about the JC itself, rather than the AM. When AM is restarted, all the containers should get restarted. I think first sentence is referring to a scenarios with Yarn High Availability. Should I just mention that JC availability is same as the AM's availability since JC resides within the AM? I can additionally mention how this will work with Yarn HA. > On Nov. 23, 2015, 7:31 p.m., Yi Pan (Data Infrastructure) wrote: > > docs/learn/documentation/versioned/container/coordinator-stream.md, line 147 > > <https://reviews.apache.org/r/40572/diff/2/?file=1135547#file1135547line147> > > > > One more benifit that would be huge to add to 5: > > - enables future design of JobCoordinator fail-over since it serves as > > the single source of truth of current Job state. Are you referring to the failover of the Job Coordinator or the failover of containers?? - Navina ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/40572/#review107623 ----------------------------------------------------------- On Nov. 23, 2015, 3:38 a.m., Navina Ramesh wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/40572/ > ----------------------------------------------------------- > > (Updated Nov. 23, 2015, 3:38 a.m.) > > > Review request for samza, Boris Shkolnik, Yan Fang, Chris Riccomini, Jagadish > Venkatraman, Xinyu Liu, and Yi Pan (Data Infrastructure). > > > Bugs: SAMZA-614 > https://issues.apache.org/jira/browse/SAMZA-614 > > > Repository: samza > > > Description > ------- > > SAMZA-614 - Document coordinator stream > Updated configuration-table.html > > > Diffs > ----- > > docs/learn/documentation/versioned/container/coordinator-stream.md > PRE-CREATION > docs/learn/documentation/versioned/container/serialization.md > 6bcb641f2e3d168581274e468746f9f3078d4b8a > docs/learn/documentation/versioned/index.html > dec5be15453a8e5794a9c66b0010803f591ed0cf > docs/learn/documentation/versioned/jobs/configuration-table.html > b5d3813ab44062f61cb33e4b37bb7548d8ff0617 > > Diff: https://reviews.apache.org/r/40572/diff/ > > > Testing > ------- > > Locally tested the site > > > Thanks, > > Navina Ramesh > >
