[
https://issues.apache.org/jira/browse/CARBONDATA-3401?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
lamber-ken updated CARBONDATA-3401:
-----------------------------------
Description:
When use sdk to write carbondata files, use +System.nanoTime()+ assign to the
start timestamp and use +System.nanoTime()+ assign to taskNo currently.
There are two bug here, first +System.nanoTime()+ is different from
+System.currentTimeMillis()+, it can only be used to measure elapsed time and
is not related to any other notion of system or wall-clock time. Second, the
carbondata file name written by sdk was different from written by spark.
was:
When use sdk to write carbondata files, use +System.nanoTime()+ assign to the
start timestamp and use +System.nanoTime()+ assign to taskNo currently.
There are two bug here, first +System.nanoTime()+ is different from
+System.currentTimeMillis()+, it can only be used to measure elapsed time and
is not related to any other notion of system or wall-clock time. Second, the
carbondata file name written by sdk was different from written by spark.+
+
> Fix the java sdk create wrong carbondata filename
> --------------------------------------------------
>
> Key: CARBONDATA-3401
> URL: https://issues.apache.org/jira/browse/CARBONDATA-3401
> Project: CarbonData
> Issue Type: Bug
> Components: data-load, other
> Affects Versions: 1.5.3
> Reporter: lamber-ken
> Priority: Major
> Fix For: 1.6.0
>
> Attachments: fix_the_java_sdk_create_wrong_carbondata_filename.patch
>
>
> When use sdk to write carbondata files, use +System.nanoTime()+ assign to the
> start timestamp and use +System.nanoTime()+ assign to taskNo currently.
> There are two bug here, first +System.nanoTime()+ is different from
> +System.currentTimeMillis()+, it can only be used to measure elapsed time and
> is not related to any other notion of system or wall-clock time. Second, the
> carbondata file name written by sdk was different from written by spark.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
