Zhijie Shen commented on YARN-3984:

[~vrushalic], thanks for picking it up. The aforementioned cases are definitely 
good to support, while the current query we want to support now (in YARN-3051 
and YARN-3049) is to retrieve all events belonging to an entity (e.g. 
application, attempt, container and etc.). With this basic query, we can easily 
distill the details that happen to the entity, such as the diagnostic msg of 
the kill event. In this case, the most efficient way is to put timestamp even 
before the event ID, so that we don't need to order the events in memory.

In addition to the key composition, I find another significant problem with the 
event store schema. If the event doesn't contain any info, it will be ignored 
then. And we cannot always guarantee user will put something into info. For 
example, user may define a KILL event without any diagnostic msg.

> Rethink event column key issue
> ------------------------------
>                 Key: YARN-3984
>                 URL: https://issues.apache.org/jira/browse/YARN-3984
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: timelineserver
>            Reporter: Zhijie Shen
>            Assignee: Vrushali C
>             Fix For: YARN-2928
> Currently, the event column key is event_id?info_key?timestamp, which is not 
> so friendly to fetching all the events of an entity and sorting them in a 
> chronologic order. IMHO, timestamp?event_id?info_key may be a better key 
> schema. I open this jira to continue the discussion about it which was 
> commented on YARN-3908.

This message was sent by Atlassian JIRA

Reply via email to