Skip to content

Conversation

@zsxwing
Copy link
Member

@zsxwing zsxwing commented Oct 3, 2014

Now the Stage page only displays "Executor"(host) for tasks. However, there may be more than one Executors running in the same host. Currently, when some task is hung, I only know the host of the faulty executor. Therefore I have to check all executors in the host.

Adding "Executor ID" in the Tasks table. would be helpful to locate the faulty executor. Here is the new page:

add_executor_id_for_tasks

@SparkQA
Copy link

SparkQA commented Oct 3, 2014

QA tests have started for PR 2642 at commit 4bbe2c7.

  • This patch merges cleanly.

@SparkQA
Copy link

SparkQA commented Oct 3, 2014

QA tests have finished for PR 2642 at commit 4bbe2c7.

  • This patch passes unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/21246/

@sryza
Copy link
Contributor

sryza commented Oct 3, 2014

As horizontal space is precious for including more metrics, might it make sense to combine "Address / Executor" and "Executor ID" into a single "Executor" column, with values like "1 / 10.37.129.2". Also, is including the port still worthwhile now that we have the ID?

@SparkQA
Copy link

SparkQA commented Oct 4, 2014

QA tests have started for PR 2642 at commit 37945af.

  • This patch merges cleanly.

@zsxwing
Copy link
Member Author

zsxwing commented Oct 4, 2014

As horizontal space is precious for including more metrics, might it make sense to combine "Address / Executor" and "Executor ID" into a single "Executor" column, with values like "1 / 10.37.129.2".

Agree. I updated to put them into one column. I use host because I feel Address should include the port number but here we only have host. The new screenshot is as follow:

executor_id_host

Also, is including the port still worthwhile now that we have the ID?

TaskInfo does not have a port field, and I cannot find an easy way to add it. However, I think Executor ID is enough. If the executor id is provided, I can use ps -ef | grep spark | grep " <executor_id> " to find the process id. Looks port cannot help me find the process id more easily.

@SparkQA
Copy link

SparkQA commented Oct 4, 2014

QA tests have finished for PR 2642 at commit 37945af.

  • This patch passes unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/21295/Test PASSed.

@andrewor14
Copy link
Contributor

LGTM. @sryza any other comments?

@sryza
Copy link
Contributor

sryza commented Oct 7, 2014

This LGTM too. Thanks for making those changes @zsxwing .

@andrewor14
Copy link
Contributor

Alright, this is going into master and 1.1

asfgit pushed a commit that referenced this pull request Oct 7, 2014
Now the Stage page only displays "Executor"(host) for tasks. However, there may be more than one Executors running in the same host. Currently, when some task is hung, I only know the host of the faulty executor. Therefore I have to check all executors in the host.

Adding "Executor ID" in the Tasks table. would be helpful to locate the faulty executor. Here is the new page:

![add_executor_id_for_tasks](https://cloud.githubusercontent.com/assets/1000778/4505774/acb9648c-4afa-11e4-8826-8768a0a60cc9.png)

Author: zsxwing <[email protected]>

Closes #2642 from zsxwing/SPARK-3777 and squashes the following commits:

37945af [zsxwing] Put Executor ID and Host into one cell
4bbe2c7 [zsxwing] [SPARK-3777] Display "Executor ID" for Tasks in Stage page

(cherry picked from commit 446063e)
Signed-off-by: Andrew Or <[email protected]>
@asfgit asfgit closed this in 446063e Oct 7, 2014
@zsxwing zsxwing deleted the SPARK-3777 branch October 8, 2014 03:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants