You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by yunjzhang <gi...@git.apache.org> on 2018/08/21 23:29:21 UTC
[GitHub] spark pull request #22177: stages in wrong order within job page DAG chart
GitHub user yunjzhang opened a pull request:
https://github.com/apache/spark/pull/22177
stages in wrong order within job page DAG chart
if spark job contains multiple tasks , the order in DAG chart might be incorrect.
sample and screen snapshot can be found in jira ticket https://issues.apache.org/jira/browse/SPARK-25119
to fix this issue, just sort RDDOperationGraph Array on "task id" before UIUtils.showDagVizForJob()
## How was this patch tested?
packaging with all existing UT passed.
run complex query in Spark-sql console with the new jars
check the DAG chart in job page.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/yunjzhang/spark SPARK-25119
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/22177.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #22177
----
commit f75b3fa1a09b63617d940e779a24614b82d22905
Author: Zhang <yu...@...>
Date: 2018-08-21T23:06:08Z
stages in wrong order within job page DAG chart
----
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by yunjzhang <gi...@git.apache.org>.
Github user yunjzhang commented on the issue:
https://github.com/apache/spark/pull/22177
my fault, just fixed.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark pull request #22177: [SPARK-25119][Web UI] stages in wrong order withi...
Posted by vanzin <gi...@git.apache.org>.
Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/22177#discussion_r212134229
--- Diff: core/src/main/scala/org/apache/spark/ui/jobs/JobPage.scala ---
@@ -18,18 +18,18 @@
package org.apache.spark.ui.jobs
import java.util.Locale
+
import javax.servlet.http.HttpServletRequest
import scala.collection.mutable.{Buffer, ListBuffer}
import scala.xml.{Node, NodeSeq, Unparsed, Utility}
-
--- End diff --
+1
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: stages in wrong order within job page DAG chart
Posted by AmplabJenkins <gi...@git.apache.org>.
Github user AmplabJenkins commented on the issue:
https://github.com/apache/spark/pull/22177
Can one of the admins verify this patch?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by yunjzhang <gi...@git.apache.org>.
Github user yunjzhang commented on the issue:
https://github.com/apache/spark/pull/22177
I respect your decision.
this PR does not fix the order issue from the root.
I saw differences when jobs/tasks were shown in spark 2.1/2.3.
same insert...select was treated as 2 jobs in 2.1, while which was 1 job in 2.3.
If 2.3 can push tasks into an array in right order, extract sort is not necessary in this PR and also PR#21680
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by vanzin <gi...@git.apache.org>.
Github user vanzin commented on the issue:
https://github.com/apache/spark/pull/22177
That looks better indeed. I'll backport that one to 2.3.
@yunjzhang could you close this PR?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by vanzin <gi...@git.apache.org>.
Github user vanzin commented on the issue:
https://github.com/apache/spark/pull/22177
You're sorting on stage ID but your description says task ID. Just a small inconsistency.
This is also against branch-2.3, it should be against master.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark pull request #22177: [SPARK-25119][Web UI] stages in wrong order withi...
Posted by vanzin <gi...@git.apache.org>.
Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/22177#discussion_r212134537
--- Diff: core/src/main/scala/org/apache/spark/ui/jobs/JobPage.scala ---
@@ -337,7 +337,9 @@ private[ui] class JobPage(parent: JobsTab, store: AppStatusStore) extends WebUIP
store.executorList(false), appStartTime)
val operationGraphContent = store.asOption(store.operationGraphForJob(jobId)) match {
- case Some(operationGraph) => UIUtils.showDagVizForJob(jobId, operationGraph)
+ case Some(operationGraph) => UIUtils.showDagVizForJob(jobId, operationGraph.sortWith(
+ _.rootCluster.id.replaceAll(RDDOperationGraph.STAGE_CLUSTER_PREFIX, "").toInt
--- End diff --
+1. `replaceAll` also seems like the wrong API to use, `substring` seems more correct.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by wangyum <gi...@git.apache.org>.
Github user wangyum commented on the issue:
https://github.com/apache/spark/pull/22177
How about backport https://github.com/apache/spark/pull/21680?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25199][Web UI] stages in wrong order within job p...
Posted by MaxGekk <gi...@git.apache.org>.
Github user MaxGekk commented on the issue:
https://github.com/apache/spark/pull/22177
Probably you put wrong JIRA in the title. `SPARK-25199` -> `SPARK-25119`
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark pull request #22177: stages in wrong order within job page DAG chart
Posted by xuanyuanking <gi...@git.apache.org>.
Github user xuanyuanking commented on a diff in the pull request:
https://github.com/apache/spark/pull/22177#discussion_r212002571
--- Diff: core/src/main/scala/org/apache/spark/ui/jobs/JobPage.scala ---
@@ -18,18 +18,18 @@
package org.apache.spark.ui.jobs
import java.util.Locale
+
import javax.servlet.http.HttpServletRequest
import scala.collection.mutable.{Buffer, ListBuffer}
import scala.xml.{Node, NodeSeq, Unparsed, Utility}
-
--- End diff --
revert this changes in import.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: stages in wrong order within job page DAG chart
Posted by AmplabJenkins <gi...@git.apache.org>.
Github user AmplabJenkins commented on the issue:
https://github.com/apache/spark/pull/22177
Can one of the admins verify this patch?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark pull request #22177: [SPARK-25119][Web UI] stages in wrong order withi...
Posted by yunjzhang <gi...@git.apache.org>.
Github user yunjzhang closed the pull request at:
https://github.com/apache/spark/pull/22177
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25199][Web UI] stages in wrong order within job p...
Posted by yunjzhang <gi...@git.apache.org>.
Github user yunjzhang commented on the issue:
https://github.com/apache/spark/pull/22177
thanks for the suggestion, just rename the PR.
before fix
![image](https://user-images.githubusercontent.com/27383592/44485113-7fdbee80-a604-11e8-8587-c7f1ae559549.png)
after fix
![image](https://user-images.githubusercontent.com/27383592/44485138-92562800-a604-11e8-8694-afaee939e9d9.png)
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: stages in wrong order within job page DAG chart
Posted by AmplabJenkins <gi...@git.apache.org>.
Github user AmplabJenkins commented on the issue:
https://github.com/apache/spark/pull/22177
Can one of the admins verify this patch?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: [SPARK-25119][Web UI] stages in wrong order within job p...
Posted by vanzin <gi...@git.apache.org>.
Github user vanzin commented on the issue:
https://github.com/apache/spark/pull/22177
If you think there is still an issue, open a PR against master (not 2.3).
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark issue #22177: stages in wrong order within job page DAG chart
Posted by xuanyuanking <gi...@git.apache.org>.
Github user xuanyuanking commented on the issue:
https://github.com/apache/spark/pull/22177
Please change title to "[SPARK-25199][Web UI] XXX " as we described in http://spark.apache.org/contributing.html.
```
check the DAG chart in job page.
```
Could you also put the DAG chart screenshot after your fix?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] spark pull request #22177: stages in wrong order within job page DAG chart
Posted by xuanyuanking <gi...@git.apache.org>.
Github user xuanyuanking commented on a diff in the pull request:
https://github.com/apache/spark/pull/22177#discussion_r212003441
--- Diff: core/src/main/scala/org/apache/spark/ui/jobs/JobPage.scala ---
@@ -337,7 +337,9 @@ private[ui] class JobPage(parent: JobsTab, store: AppStatusStore) extends WebUIP
store.executorList(false), appStartTime)
val operationGraphContent = store.asOption(store.operationGraphForJob(jobId)) match {
- case Some(operationGraph) => UIUtils.showDagVizForJob(jobId, operationGraph)
+ case Some(operationGraph) => UIUtils.showDagVizForJob(jobId, operationGraph.sortWith(
+ _.rootCluster.id.replaceAll(RDDOperationGraph.STAGE_CLUSTER_PREFIX, "").toInt
--- End diff --
Add `getStageId` function in `RDDOperationGraph` to do this will be better.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org