-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/39452/#review103172
-----------------------------------------------------------


Looks pretty good. Some minor nits, should be quick to address.


docs/scheduler-http-api.md (line 27)
<https://reviews.apache.org/r/39452/#comment161142>

    1. Is it possible to link to the `SUBSCRIBE` call in the document ?
    
    2. s/The Events stream is returned/The response returned
    
    3. s/streamed in RecordIO/encoded in RecordIO



docs/scheduler-http-api.md (line 38)
<https://reviews.apache.org/r/39452/#comment161148>

    1. s/stream/response
    2. s/client/application



docs/scheduler-http-api.md (line 44)
<https://reviews.apache.org/r/39452/#comment161149>

    Nit: Can we just check for `200` here (since the response from `Subscribe` 
should be `200` for a good response ?



docs/scheduler-http-api.md (line 47)
<https://reviews.apache.org/r/39452/#comment161150>

    Nit: `Decode the response as RecordIO:`



docs/scheduler-http-api.md (line 51)
<https://reviews.apache.org/r/39452/#comment161137>

    Is `int(...)` in python unsigned and 64 bytes ? That is the requirements we 
mention in our docs.



docs/scheduler-http-api.md (line 59)
<https://reviews.apache.org/r/39452/#comment161141>

    Should we also mention why just encoding one event per chunk won't suffice 
and why we needed this ?
    
    How about:
    
    ```Network intermediataries e.g. proxies are free to change the chunk 
boundaries and this should not have any effect on the recipient 
application(scheduler layer). We wanted a way to delimit/encode two events for 
JSON/Protobuf responses consistently and RecordIO format allowed us to do 
that.```


- Anand Mazumdar


On Oct. 19, 2015, 9:07 p.m., Marco Massenzio wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/39452/
> -----------------------------------------------------------
> 
> (Updated Oct. 19, 2015, 9:07 p.m.)
> 
> 
> Review request for mesos, Anand Mazumdar and Vinod Kone.
> 
> 
> Bugs: MESOS-3566
>     https://issues.apache.org/jira/browse/MESOS-3566
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Added the description of the RecordIO format to the HTTP API
> document with example code (Python) to decode.
> 
> 
> Diffs
> -----
> 
>   docs/scheduler-http-api.md de6cfc9e009a857ca45291b2dadce2a3b8199787 
> 
> Diff: https://reviews.apache.org/r/39452/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Marco Massenzio
> 
>

Reply via email to