[jira] [Commented] (SPARK-6752) Allow StreamingContext to be recreated from checkpoint and existing SparkContext
[
https://issues.apache.org/jira/browse/SPARK-6752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14540704#comment-14540704
]
Apache Spark commented on SPARK-6752:
-
User 'tdas' has created a pull request for this issue:
https://github.com/apache/spark/pull/6096
> Allow StreamingContext to be recreated from checkpoint and existing
> SparkContext
>
>
> Key: SPARK-6752
> URL: https://issues.apache.org/jira/browse/SPARK-6752
> Project: Spark
> Issue Type: Improvement
> Components: Streaming
>Affects Versions: 1.1.1, 1.2.1, 1.3.1
>Reporter: Tathagata Das
>Assignee: Tathagata Das
>Priority: Critical
> Fix For: 1.4.0
>
>
> Currently if you want to create a StreamingContext from checkpoint
> information, the system will create a new SparkContext. This prevent
> StreamingContext to be recreated from checkpoints in managed environments
> where SparkContext is precreated.
> Proposed solution: Introduce the following methods on StreamingContext
> 1. {{new StreamingContext(checkpointDirectory, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> 2. {{new StreamingContext(checkpointDirectory, hadoopConf, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> and hadoop conf to read the checkpoint
> 3. {{StreamingContext.getOrCreate(checkpointDirectory, sparkContext,
> createFunction: SparkContext => StreamingContext)}}
> - If checkpoint file exists, then recreate StreamingContext using the
> provided SparkContext (that is, call 1.), else create StreamingContext using
> the provided createFunction
> Also, the corresponding Java and Python API has to be added as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-6752) Allow StreamingContext to be recreated from checkpoint and existing SparkContext
[
https://issues.apache.org/jira/browse/SPARK-6752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14540679#comment-14540679
]
Tathagata Das commented on SPARK-6752:
--
I am revising this feature. The earlier version passed the active SparkContext
explicitly through a new set of Java and Scala API. The drawbacks are.
1. Hard to implement in python.
2. New API introduced. This is even more confusing since we are introducing
{{getActiveOrCreate}} in SPARK-7553
Furthermore, there is now a direct way get an existing active SparkContext or
create a new on - {{SparkContext.getOrCreate(conf)}}. Its better to use this to
get the SparkContext rather than have a new API to explicitly pass the context.
> Allow StreamingContext to be recreated from checkpoint and existing
> SparkContext
>
>
> Key: SPARK-6752
> URL: https://issues.apache.org/jira/browse/SPARK-6752
> Project: Spark
> Issue Type: Improvement
> Components: Streaming
>Affects Versions: 1.1.1, 1.2.1, 1.3.1
>Reporter: Tathagata Das
>Assignee: Tathagata Das
>Priority: Critical
> Fix For: 1.4.0
>
>
> Currently if you want to create a StreamingContext from checkpoint
> information, the system will create a new SparkContext. This prevent
> StreamingContext to be recreated from checkpoints in managed environments
> where SparkContext is precreated.
> Proposed solution: Introduce the following methods on StreamingContext
> 1. {{new StreamingContext(checkpointDirectory, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> 2. {{new StreamingContext(checkpointDirectory, hadoopConf, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> and hadoop conf to read the checkpoint
> 3. {{StreamingContext.getOrCreate(checkpointDirectory, sparkContext,
> createFunction: SparkContext => StreamingContext)}}
> - If checkpoint file exists, then recreate StreamingContext using the
> provided SparkContext (that is, call 1.), else create StreamingContext using
> the provided createFunction
> Also, the corresponding Java and Python API has to be added as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-6752) Allow StreamingContext to be recreated from checkpoint and existing SparkContext
[
https://issues.apache.org/jira/browse/SPARK-6752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14518927#comment-14518927
]
Apache Spark commented on SPARK-6752:
-
User 'tdas' has created a pull request for this issue:
https://github.com/apache/spark/pull/5773
> Allow StreamingContext to be recreated from checkpoint and existing
> SparkContext
>
>
> Key: SPARK-6752
> URL: https://issues.apache.org/jira/browse/SPARK-6752
> Project: Spark
> Issue Type: Improvement
> Components: Streaming
>Affects Versions: 1.1.1, 1.2.1, 1.3.1
>Reporter: Tathagata Das
>Assignee: Tathagata Das
>Priority: Critical
> Fix For: 1.4.0
>
>
> Currently if you want to create a StreamingContext from checkpoint
> information, the system will create a new SparkContext. This prevent
> StreamingContext to be recreated from checkpoints in managed environments
> where SparkContext is precreated.
> Proposed solution: Introduce the following methods on StreamingContext
> 1. {{new StreamingContext(checkpointDirectory, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> 2. {{new StreamingContext(checkpointDirectory, hadoopConf, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> and hadoop conf to read the checkpoint
> 3. {{StreamingContext.getOrCreate(checkpointDirectory, sparkContext,
> createFunction: SparkContext => StreamingContext)}}
> - If checkpoint file exists, then recreate StreamingContext using the
> provided SparkContext (that is, call 1.), else create StreamingContext using
> the provided createFunction
> Also, the corresponding Java and Python API has to be added as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-6752) Allow StreamingContext to be recreated from checkpoint and existing SparkContext
[
https://issues.apache.org/jira/browse/SPARK-6752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14512620#comment-14512620
]
Sean Owen commented on SPARK-6752:
--
It's an issue with the version of Commons Lang (2.x). Just use
{{AtomicBoolean}} here and avoid the library altogether
> Allow StreamingContext to be recreated from checkpoint and existing
> SparkContext
>
>
> Key: SPARK-6752
> URL: https://issues.apache.org/jira/browse/SPARK-6752
> Project: Spark
> Issue Type: Improvement
> Components: Streaming
>Affects Versions: 1.1.1, 1.2.1, 1.3.1
>Reporter: Tathagata Das
>Assignee: Tathagata Das
>Priority: Critical
> Fix For: 1.4.0
>
>
> Currently if you want to create a StreamingContext from checkpoint
> information, the system will create a new SparkContext. This prevent
> StreamingContext to be recreated from checkpoints in managed environments
> where SparkContext is precreated.
> Proposed solution: Introduce the following methods on StreamingContext
> 1. {{new StreamingContext(checkpointDirectory, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> 2. {{new StreamingContext(checkpointDirectory, hadoopConf, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> and hadoop conf to read the checkpoint
> 3. {{StreamingContext.getOrCreate(checkpointDirectory, sparkContext,
> createFunction: SparkContext => StreamingContext)}}
> - If checkpoint file exists, then recreate StreamingContext using the
> provided SparkContext (that is, call 1.), else create StreamingContext using
> the provided createFunction
> Also, the corresponding Java and Python API has to be added as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-6752) Allow StreamingContext to be recreated from checkpoint and existing SparkContext
[
https://issues.apache.org/jira/browse/SPARK-6752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14486228#comment-14486228
]
Apache Spark commented on SPARK-6752:
-
User 'tdas' has created a pull request for this issue:
https://github.com/apache/spark/pull/5428
> Allow StreamingContext to be recreated from checkpoint and existing
> SparkContext
>
>
> Key: SPARK-6752
> URL: https://issues.apache.org/jira/browse/SPARK-6752
> Project: Spark
> Issue Type: Improvement
> Components: Streaming
>Affects Versions: 1.1.1, 1.2.1, 1.3.1
>Reporter: Tathagata Das
>Assignee: Tathagata Das
>Priority: Critical
>
> Currently if you want to create a StreamingContext from checkpoint
> information, the system will create a new SparkContext. This prevent
> StreamingContext to be recreated from checkpoints in managed environments
> where SparkContext is precreated.
> Proposed solution: Introduce the following methods on StreamingContext
> 1. {{new StreamingContext(checkpointDirectory, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> 2. {{new StreamingContext(checkpointDirectory, hadoopConf, sparkContext)}}
> - Recreate StreamingContext from checkpoint using the provided SparkContext
> and hadoop conf to read the checkpoint
> 3. {{StreamingContext.getOrCreate(checkpointDirectory, sparkContext,
> createFunction: SparkContext => StreamingContext)}}
> - If checkpoint file exists, then recreate StreamingContext using the
> provided SparkContext (that is, call 1.), else create StreamingContext using
> the provided createFunction
> Also, the corresponding Java and Python API has to be added as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
