samza-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Cameron Lee (Jira)" <j...@apache.org>
Subject [jira] [Updated] (SAMZA-2304) Existing container locality mapping is incorrect when building job model
Date Tue, 20 Aug 2019 17:06:00 GMT

     [ https://issues.apache.org/jira/browse/SAMZA-2304?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Cameron Lee updated SAMZA-2304:
-------------------------------
    Affects Version/s: 1.2
                       1.1
                       1.0
          Description: When building the job model, the existing container locality is read,
so that the task to container grouping can be calculated. However, the existing container
locality mapping is incorrect, in that it always has one extra container than the job container
count. If a grouper (e.g. GroupByContainerIds) uses the mapping to figure out the container
models, then it might return the incorrect number of containers for the job model.  (was:
When building the job model, the existing container locality is read, so that the task to
container grouping can be calculated. However, the existing container locality mapping is
incorrect, in that it always has one extra container than the job container count. If a grouper
uses the mapping to figure out the container models, then it might return the incorrect number
of containers for the job model.)

> Existing container locality mapping is incorrect when building job model
> ------------------------------------------------------------------------
>
>                 Key: SAMZA-2304
>                 URL: https://issues.apache.org/jira/browse/SAMZA-2304
>             Project: Samza
>          Issue Type: Bug
>    Affects Versions: 1.0, 1.1, 1.2
>            Reporter: Cameron Lee
>            Assignee: Cameron Lee
>            Priority: Major
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> When building the job model, the existing container locality is read, so that the task
to container grouping can be calculated. However, the existing container locality mapping
is incorrect, in that it always has one extra container than the job container count. If a
grouper (e.g. GroupByContainerIds) uses the mapping to figure out the container models, then
it might return the incorrect number of containers for the job model.



--
This message was sent by Atlassian Jira
(v8.3.2#803003)

Mime
View raw message