[jira] [Commented] (HIVE-12753) Improve performance of LazyBinaryColumnarSerDe for Text strings
[
https://issues.apache.org/jira/browse/HIVE-12753?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15082545#comment-15082545
]
Heng Chen commented on HIVE-12753:
--
What's the meaning of the failed testcase? relates with this patch?
> Improve performance of LazyBinaryColumnarSerDe for Text strings
> ---
>
> Key: HIVE-12753
> URL: https://issues.apache.org/jira/browse/HIVE-12753
> Project: Hive
> Issue Type: Bug
> Components: Serializers/Deserializers
>Affects Versions: 1.2.1
>Reporter: David Phillips
> Attachments: HIVE-12753.patch
>
>
> {{LazyBinaryColumnarSerDe}} needs to special case writing empty strings, but
> it does so by calling {{getPrimitiveJavaObject()}} rather than
> {{getPrimitiveWritableObject()}} on {{StringObjectInspector}}. If the object
> is {{Text}}, then this requires a conversion to a Java {{String}} just to get
> the length.
> Unfortunately, the best method to call depends on the object type (or the
> object inspector type). One solution is to call
> {{getPrimitiveWritableObject()}} if the object is an instance of {{Text}} or
> {{BytesWritable}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Updated] (HIVE-12753) Improve performance of LazyBinaryColumnarSerDe for Text strings
[
https://issues.apache.org/jira/browse/HIVE-12753?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Heng Chen updated HIVE-12753:
-
Attachment: HIVE-12753.patch
Make a simple patch on master. :)
> Improve performance of LazyBinaryColumnarSerDe for Text strings
> ---
>
> Key: HIVE-12753
> URL: https://issues.apache.org/jira/browse/HIVE-12753
> Project: Hive
> Issue Type: Bug
> Components: Serializers/Deserializers
>Affects Versions: 1.2.1
>Reporter: David Phillips
> Attachments: HIVE-12753.patch
>
>
> {{LazyBinaryColumnarSerDe}} needs to special case writing empty strings, but
> it does so by calling {{getPrimitiveJavaObject()}} rather than
> {{getPrimitiveWritableObject()}} on {{StringObjectInspector}}. If the object
> is {{Text}}, then this requires a conversion to a Java {{String}} just to get
> the length.
> Unfortunately, the best method to call depends on the object type (or the
> object inspector type). One solution is to call
> {{getPrimitiveWritableObject()}} if the object is an instance of {{Text}} or
> {{BytesWritable}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Updated] (HIVE-12665) Hive Server2 NOT work after running some time.
[
https://issues.apache.org/jira/browse/HIVE-12665?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Heng Chen updated HIVE-12665:
-
Description:
I use jdbc in my project to submit job.
After running some days, Hive Server2 will not work any more. Every time i
submit a job, it returns error like below
{code}
Error while processing statement: FAILED: Execution Error, return code 1 from
org.apache.hadoop.hive.ql.exec.mr.MapRedTask
{code}
It seems MR Server thought MR client is stopped, so relates MR job was
killed.
After i restart Hive Server2, everything will be OK.
was:
I use jdbc in my project to submit job.
After running some days, Hive Thrift Server will not work any more. Every
time i submit a job, it returns error like below
{code}
Error while processing statement: FAILED: Execution Error, return code 1 from
org.apache.hadoop.hive.ql.exec.mr.MapRedTask
{code}
It seems MR Server thought MR client is stopped, so relates MR job was
killed.
After i restart Hive Thrift Server, everything will be OK.
> Hive Server2 NOT work after running some time.
> --
>
> Key: HIVE-12665
> URL: https://issues.apache.org/jira/browse/HIVE-12665
> Project: Hive
> Issue Type: Bug
>Affects Versions: 0.14.0
>Reporter: Heng Chen
>Priority: Critical
>
> I use jdbc in my project to submit job.
> After running some days, Hive Server2 will not work any more. Every time i
> submit a job, it returns error like below
> {code}
> Error while processing statement: FAILED: Execution Error, return code 1 from
> org.apache.hadoop.hive.ql.exec.mr.MapRedTask
> {code}
> It seems MR Server thought MR client is stopped, so relates MR job was
> killed.
> After i restart Hive Server2, everything will be OK.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Updated] (HIVE-12665) Hive Server2 NOT work after running some time.
[
https://issues.apache.org/jira/browse/HIVE-12665?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Heng Chen updated HIVE-12665:
-
Summary: Hive Server2 NOT work after running some time. (was: Hive Thrift
Server NOT work after running some time.)
> Hive Server2 NOT work after running some time.
> --
>
> Key: HIVE-12665
> URL: https://issues.apache.org/jira/browse/HIVE-12665
> Project: Hive
> Issue Type: Bug
>Affects Versions: 0.14.0
>Reporter: Heng Chen
>Priority: Critical
>
> I use jdbc in my project to submit job.
> After running some days, Hive Thrift Server will not work any more. Every
> time i submit a job, it returns error like below
> {code}
> Error while processing statement: FAILED: Execution Error, return code 1 from
> org.apache.hadoop.hive.ql.exec.mr.MapRedTask
> {code}
> It seems MR Server thought MR client is stopped, so relates MR job was
> killed.
> After i restart Hive Thrift Server, everything will be OK.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
