[
https://issues.apache.org/jira/browse/YARN-7100?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Eric Yang updated YARN-7100:
----------------------------
Attachment: YARN-7100.004.yarn-native-services.patch
Rebase patch based on the current HEAD of yarn-native-services branch.
> YARN service api can not reuse json file serialized in hdfs
> -----------------------------------------------------------
>
> Key: YARN-7100
> URL: https://issues.apache.org/jira/browse/YARN-7100
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: yarn-native-services
> Affects Versions: 3.0.0-beta1
> Reporter: Eric Yang
> Attachments: YARN-7100.001.patch,
> YARN-7100.002.yarn-native-services.patch,
> YARN-7100.003.yarn-native-services.patch,
> YARN-7100.004.yarn-native-services.patch
>
>
> org.apache.hadoop.yarn.service.api.records.Resource has a new method
> introduced in YARN-6903 for casting memory from string to long value.
> However, the method name getMemoryMB introduces new output in json that looks
> like this:
> {code}
> "resource" : {
> "uri" : null,
> "profile" : null,
> "cpus" : 1,
> "memory" : "2048",
> "memory_mb" : 2048
> },
> {code}
> This prevents the file to be resubmitted to services api because memory_mb
> property is unknown to the REST API. It may be better to rename getMemoryMB
> method to calcMemoryMB to avoid the method to be serialized unintentionally.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
