You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2020/09/30 08:30:18 UTC

[GitHub] [spark] zhli1142015 opened a new pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

zhli1142015 opened a new pull request #29908:
URL: https://github.com/apache/spark/pull/29908


   <!--
   Thanks for sending a pull request!  Here are some tips for you:
     1. If this is your first time, please read our contributor guidelines: https://spark.apache.org/contributing.html
     2. Ensure you have added or run the appropriate tests for your PR: https://spark.apache.org/developer-tools.html
     3. If the PR is unfinished, add '[WIP]' in your PR title, e.g., '[WIP][SPARK-XXXX] Your PR title ...'.
     4. Be sure to keep the PR description updated to reflect all changes.
     5. Please write your PR title to summarize what this PR proposes.
     6. If possible, provide a concise example to reproduce the issue for a faster review.
     7. If you want to add a new configuration, please read the guideline first for naming configurations in
        'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
   -->
   
   ### What changes were proposed in this pull request?
   <!--
   Please clarify what changes you are proposing. The purpose of this section is to outline the changes and how this PR fixes the issue. 
   If possible, please consider writing useful notes for better and faster reviews in your PR. See the examples below.
     1. If you refactor some codes with changing classes, showing the class hierarchy will help reviewers.
     2. If you fix some SQL features, you can provide some references of other DBMSes.
     3. If there is design documentation, please add the link.
     4. If there is a discussion in the mailing list, please add the link.
   -->
   
   
   ### Why are the changes needed?
   <!--
   Please clarify why the changes are needed. For instance,
     1. If you propose a new API, clarify the use case for a new API.
     2. If you fix a bug, you can clarify why it is a bug.
   -->
   Event log contains all tasks' metrics data, which are useful for performance debugging. By now, spark UI only displays final aggregation results, much information is hidden by this way. If spark UI could provide time series data view, it would be more helpful to performance debugging problems. We would like to build application statistics page in history server based on task metrics to provide more straight forward insight for spark application.
   Below are views in application statistics page:
   **Cluster Throughput**: sum of completed tasks, stages, and jobs per minute. (associated stage Ids can be viewed in tool tip message.)
   ![cluster throughput](https://user-images.githubusercontent.com/10524738/94659549-526ccf00-02b9-11eb-824c-75f55f3fb728.JPG)
   **IO (Shuffle and Read) Data**: sum of total shuffle read bytes, shuffle write bytes and read bytes per minute.
   ![IOData](https://user-images.githubusercontent.com/10524738/94659699-8a741200-02b9-11eb-80e6-9271828fc3b1.JPG)
   **Task Duration Time**: application level 50% and 90% percentile of task duration per minute.
   ![taskduration](https://user-images.githubusercontent.com/10524738/94659768-a37cc300-02b9-11eb-8147-c6ff1c2dfc1f.JPG)
   **Task Duration Time Component %**: percentage of scheduler delay, computing time, shuffle read, task deserialization, result serialization, shuffle write, get result time by task duration per minute.
   ![taskdurationpercentile](https://user-images.githubusercontent.com/10524738/94659888-cd35ea00-02b9-11eb-8c82-5ddf26a7c9ff.JPG)
   **Task JVM GC Time %**: percentage of task JVM GC time by task duration per minute.
   ![taskgcpercentile](https://user-images.githubusercontent.com/10524738/94659984-f48cb700-02b9-11eb-819a-d35f3529f68c.JPG)
   
   Application statistics page is only available in spark history server. Aggregated data is generated when parsing event log file and store in KVStore. Metrics data is aggregated to one data instance per minute (based on task finish time). For example, if task a finish time is in (t1 - 1minute, t1],a's data is added to data instance t1. This follows same approach of executors metrics.
   From my test there is no much increasing for kv store size and replaying time. Here is my local test result. Impact to replay time may be little different for different applications, but it should be too big.
   ![data](https://user-images.githubusercontent.com/10524738/94661444-ea6bb800-02bb-11eb-8951-032c4df8f2c5.JPG)
   
   
   ### Does this PR introduce _any_ user-facing change?
   <!--
   Note that it means *any* user-facing change including all aspects such as the documentation fix.
   If yes, please clarify the previous behavior and the change this PR proposes - provide the console output, description and/or an example to show the behavior difference if possible.
   If possible, please also clarify if this is a user-facing change compared to the released Spark versions or within the unreleased branches such as master.
   If no, write 'No'.
   -->
   User facing change compared to master: Add application statistics page under jobs tab and new page link in jobs page.
   Entry point:
   ![entrypoint](https://user-images.githubusercontent.com/10524738/94660218-40d7f700-02ba-11eb-878e-c02de9e11a9d.JPG)
   
   Application statistics page:
   ![screencapture-localhost-19090-history-application-1598927142852-0002-1-jobs-statistics-2020-09-30-00_30_48](https://user-images.githubusercontent.com/10524738/94658260-6e6f7100-02b7-11eb-9bba-8ee4ef09abc3.png)
   
   ### How was this patch tested?
   <!--
   If tests were added, say they were added here. Please make sure to add some test cases that check the changes thoroughly including negative and positive cases if possible.
   If it was tested in a way different from regular unit tests, please clarify how you tested step by step, ideally copy and paste-able, so that other reviewers can test and check, and descendants can verify in the future.
   If tests were not added, please describe why they were not added and/or why it was difficult to add.
   -->
   1. manual test
   2. add new unit test


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] zhli1142015 edited a comment on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
zhli1142015 edited a comment on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701833414


   @tgravescs , thanks for your comments. The views here are all application level. You are right, "cluster throughput" is confusing, I use this as we run spark per cluster per application in our cases, but it's not suitable for other users. i think "Execution Throughput" here is better. 
   
   if tasks start and finish within a minute, it's still shown here. for example task A started and finished within (t1 - 1 min, t1], it's value is recorded in data point t1. 
   
   For long application, the default time range may be hours, all the graphs support zooming in so we can select small time rang for details (but minimum interval is not changed). for shorter application, we can use smaller aggregation period (this is a config for all applications rendered in HS), i also added config for this, for example "5 seconds" as below. 
   ![5seconds](https://user-images.githubusercontent.com/10524738/94761479-bbe2ff80-03d7-11eb-92f4-0926c79d9f78.JPG)
   
   Original data is from tasks metrics in event log, views above are based on my experience about using spark, welcome to any suggestions.  If any other valuable view or aggregation data / dimension in your mind, we can talk about to see if it's easy to implement.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] AmplabJenkins commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
AmplabJenkins commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701246966


   Can one of the admins verify this patch?


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] AmplabJenkins removed a comment on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
AmplabJenkins removed a comment on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701246353


   Can one of the admins verify this patch?


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] zhli1142015 commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
zhli1142015 commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701833414


   @tgravescs , thanks for your comments. The views here are all application level. You are right, "cluster throughput" is confusing, I use it as we runs spark per yarn cluster per application in our cases, but it's not suitable for other users. i think "Execution Throughput" here is better. 
   
   if tasks start and finish within a minute, it's still shown here. for example task A started and finished within (t1 - 1 min, t1], it's value is recorded in data point t1. 
   
   For long application, the default time range may be hours, all the graphs support zooming in so we can select small time rang for details (but minimum interval is not changed). for shorter application, we can use smaller aggregation period, i also added config for this, for example "5 seconds" as below. 
   ![5seconds](https://user-images.githubusercontent.com/10524738/94761479-bbe2ff80-03d7-11eb-92f4-0926c79d9f78.JPG)
   
   Original data is from tasks metrics in event log, views above are based on my experience about using spark, welcome to any suggestions.  If any other valuable view or aggregation data / dimension in your mind, we can talk about to see if it's easy to implement.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] github-actions[bot] closed pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
github-actions[bot] closed pull request #29908:
URL: https://github.com/apache/spark/pull/29908


   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] github-actions[bot] commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
github-actions[bot] commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-757394266


   We're closing this PR because it hasn't been updated in a while. This isn't a judgement on the merit of the PR in any way. It's just a way of keeping the PR queue manageable.
   If you'd like to revive this PR, please reopen it and ask a committer to remove the Stale tag!


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] AmplabJenkins commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
AmplabJenkins commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701246353


   Can one of the admins verify this patch?


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] tgravescs commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
tgravescs commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-701469354


   thanks for the changes more ui changes to help debugging are great. I've been wanting something that does more timeline view to be able to see things like active executors as well as things like cached data throughout the application time since that goes away as well.
   
   the proposal here is the stats are still per application, correct?  Not combining applications.
   Assuming its within an application..
   A few high level comments just looking at screen shots.  "Cluster Throughput" might be confusing if inside an application.  Perhaps something more like Execution activity or  execution throughput.  your description says " sum of completed tasks, stages, and jobs per minute."   So if tasks start and finish within a minute its still shown here?   this view is over hours long I would be curious to see what this looks like for shorter applications.  I would also rather see these as a twisty like the event timeline on the page and not be the first thing the user sees.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


[GitHub] [spark] github-actions[bot] commented on pull request #29908: [SPARK-33033][WEBUI] display time series view for task metrics in history server

Posted by GitBox <gi...@apache.org>.
github-actions[bot] commented on pull request #29908:
URL: https://github.com/apache/spark/pull/29908#issuecomment-757394266


   We're closing this PR because it hasn't been updated in a while. This isn't a judgement on the merit of the PR in any way. It's just a way of keeping the PR queue manageable.
   If you'd like to revive this PR, please reopen it and ask a committer to remove the Stale tag!


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org