You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2022/04/12 15:37:18 UTC

[GitHub] [flink] pnowojski opened a new pull request, #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

pnowojski opened a new pull request, #19445:
URL: https://github.com/apache/flink/pull/19445

   Additionally mark TaskExecutorRecoveryTest as an ITCase due to memory consumption and starting up TaskManagers
   
   ## What is the purpose of the change
   
   Purpose of this change is to decrease build times via increasing forkCounts.
   
   ## Verifying this change
   
   Covered by existing test base.
   
   ## Does this pull request potentially affect one of the following parts:
   
     - Dependencies (does it add or upgrade a dependency): (yes / **no**)
     - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: (yes / **no**)
     - The serializers: (yes / **no** / don't know)
     - The runtime per-record code paths (performance sensitive): (yes / **no** / don't know)
     - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: (yes / **no** / don't know)
     - The S3 file system connector: (yes / **no** / don't know)
   
   ## Documentation
   
     - Does this pull request introduce a new feature? (yes / **no**)
     - If yes, how is the feature documented? (**not applicable** / docs / JavaDocs / not documented)
   


-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r855835204


##########
flink-runtime/src/test/java/org/apache/flink/runtime/taskexecutor/TaskExecutorRecoveryITCase.java:
##########
@@ -58,7 +58,7 @@
 
 /** Recovery tests for {@link TaskExecutor}. */
 @ExtendWith(TestLoggerExtension.class)
-class TaskExecutorRecoveryTest {
+class TaskExecutorRecoveryITCase {

Review Comment:
   `Could not allocate enough memory segments for NetworkBufferPool (required (Mb): 1024, allocated (Mb): 1022, missing (Mb): 2)`
   
   Looks like we're using NETWORK_MEMORY_MAX (1g) by default, which pushes us a _bit_ over the memory budget when starting 2 TMs...



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850197802


##########
pom.xml:
##########
@@ -95,13 +95,18 @@ under the License.
 		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 		<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 		<hadoop.version>2.8.5</hadoop.version>
+		<flink.XmxITCase>2048m</flink.XmxITCase>
+		<flink.XmxUnitTest>1024m</flink.XmxUnitTest>
 		<!-- Need to use a user property here because the surefire
 			 forkCount is not exposed as a property. With this we can set
 			 it on the "mvn" commandline in travis. -->
-		<flink.forkCount>1C</flink.forkCount>
+		<!-- Number of forkCounts for ITCase and UnitTest should take into account allocated memory
+			 to the jvm (-Xmx) and the available memory on the machine running the test -->
+		<flink.forkCountITCase>1C</flink.forkCountITCase>
+		<flink.forkCountUnitTest>2C</flink.forkCountUnitTest>

Review Comment:
   I'm fine with that, I also wasn't sure about the 1C/2C setting in the first place. Will do it in a separate commit.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850197001


##########
pom.xml:
##########
@@ -95,13 +95,18 @@ under the License.
 		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 		<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 		<hadoop.version>2.8.5</hadoop.version>
+		<flink.XmxITCase>2048m</flink.XmxITCase>
+		<flink.XmxUnitTest>1024m</flink.XmxUnitTest>

Review Comment:
   My intention here was to have the memory and forkCount defined in one place, close to one another, as they are very tightly connected, to make this connection more obvious and easier to spot. So I would prefer to keep it as it is here. Also this way we reduce a chance that someone will copy out those hardcoded values somewhere else without thinking everything through introducing some subtle intermittent problems with build instability.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850371496


##########
flink-end-to-end-tests/run-nightly-tests.sh:
##########
@@ -258,7 +258,7 @@ function run_group_2 {
     LOG4J_PROPERTIES=${END_TO_END_DIR}/../tools/ci/log4j.properties
 
     MVN_LOGGING_OPTIONS="-Dlog.dir=${DEBUG_FILES_OUTPUT_DIR} -DlogBackupDir=${DEBUG_FILES_OUTPUT_DIR} -Dlog4j.configurationFile=file://$LOG4J_PROPERTIES"
-    MVN_COMMON_OPTIONS="-Dflink.forkCount=2 -Dfast -Pskip-webui-build"
+    MVN_COMMON_OPTIONS="-Dflink.forkCountITCase=2 -Dflink.forkCountUnitTest=4 -Dfast -Pskip-webui-build"

Review Comment:
   Ops, good point. Sorry about that.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850336421


##########
flink-runtime/src/test/java/org/apache/flink/runtime/taskexecutor/TaskExecutorRecoveryITCase.java:
##########
@@ -58,7 +58,7 @@
 
 /** Recovery tests for {@link TaskExecutor}. */
 @ExtendWith(TestLoggerExtension.class)
-class TaskExecutorRecoveryTest {
+class TaskExecutorRecoveryITCase {

Review Comment:
   Without this change, two out of two azure runs failed with OOM when running this test. Dunno why.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r849239021


##########
pom.xml:
##########
@@ -1562,7 +1567,6 @@ under the License.
 				<configuration>
 					<!-- enables TCP/IP communication between surefire and forked JVM-->
 					<forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
-					<forkCount>${flink.forkCount}</forkCount>

Review Comment:
   there are a number of modules that override the forkCount to 1 using the general forkCount option (e.g., various connector modules), which afaik would now always ignored because the execution-specific option takes precedence.
   
   I don't know whether that's a problem though. I haven't seen an increase in test instabilities when I used these commits in a WIP branch, but those also didn't ran on Azure. (but in my branch we only changed the unit test fork count)
   In any case we either need to remove these now-ignored forkCount settings, or adjust the poms to override it in the execution instead (or modify the properties; I think that works as well).
   I would be inclined to just remove them and see what happens.
   
   huh...there should also be some forkReuse settings that are now ignored after fee8217a65bdbcf2eed8b2c8a31852f84f1022ad. Might wanna get rid of those as well...



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r849365534


##########
pom.xml:
##########
@@ -1562,7 +1567,6 @@ under the License.
 				<configuration>
 					<!-- enables TCP/IP communication between surefire and forked JVM-->
 					<forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
-					<forkCount>${flink.forkCount}</forkCount>

Review Comment:
   I'm looking into preparing a cleanup PR that removes reuseForks, other custom forkCounts etc.
   Then we should also get a clearer picture of how much this change  specifically gets us.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850371496


##########
flink-end-to-end-tests/run-nightly-tests.sh:
##########
@@ -258,7 +258,7 @@ function run_group_2 {
     LOG4J_PROPERTIES=${END_TO_END_DIR}/../tools/ci/log4j.properties
 
     MVN_LOGGING_OPTIONS="-Dlog.dir=${DEBUG_FILES_OUTPUT_DIR} -DlogBackupDir=${DEBUG_FILES_OUTPUT_DIR} -Dlog4j.configurationFile=file://$LOG4J_PROPERTIES"
-    MVN_COMMON_OPTIONS="-Dflink.forkCount=2 -Dfast -Pskip-webui-build"
+    MVN_COMMON_OPTIONS="-Dflink.forkCountITCase=2 -Dflink.forkCountUnitTest=4 -Dfast -Pskip-webui-build"

Review Comment:
   Ops, good point. Sorry about that. Fixed, but it made commits history a bit whacky (2nd commit removing what have just 1st commit added)



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] flinkbot commented on pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
flinkbot commented on PR #19445:
URL: https://github.com/apache/flink/pull/19445#issuecomment-1096894390

   <!--
   Meta data
   {
     "version" : 1,
     "metaDataEntries" : [ {
       "hash" : "e7b0dc90e48047af924a07d71f255ce7bfb52fc7",
       "status" : "UNKNOWN",
       "url" : "TBD",
       "triggerID" : "e7b0dc90e48047af924a07d71f255ce7bfb52fc7",
       "triggerType" : "PUSH"
     } ]
   }-->
   ## CI report:
   
   * e7b0dc90e48047af924a07d71f255ce7bfb52fc7 UNKNOWN
   
   <details>
   <summary>Bot commands</summary>
     The @flinkbot bot supports the following commands:
   
    - `@flinkbot run azure` re-run the last Azure build
   </details>


-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol merged pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol merged PR #19445:
URL: https://github.com/apache/flink/pull/19445


-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850344087


##########
flink-end-to-end-tests/run-nightly-tests.sh:
##########
@@ -258,7 +258,7 @@ function run_group_2 {
     LOG4J_PROPERTIES=${END_TO_END_DIR}/../tools/ci/log4j.properties
 
     MVN_LOGGING_OPTIONS="-Dlog.dir=${DEBUG_FILES_OUTPUT_DIR} -DlogBackupDir=${DEBUG_FILES_OUTPUT_DIR} -Dlog4j.configurationFile=file://$LOG4J_PROPERTIES"
-    MVN_COMMON_OPTIONS="-Dflink.forkCount=2 -Dfast -Pskip-webui-build"
+    MVN_COMMON_OPTIONS="-Dflink.forkCountITCase=2 -Dflink.forkCountUnitTest=4 -Dfast -Pskip-webui-build"

Review Comment:
   I don't think we need to set these explicitly anymore.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r849239021


##########
pom.xml:
##########
@@ -1562,7 +1567,6 @@ under the License.
 				<configuration>
 					<!-- enables TCP/IP communication between surefire and forked JVM-->
 					<forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
-					<forkCount>${flink.forkCount}</forkCount>

Review Comment:
   there are a number of modules that override the forkCount to 1 using the general forkCount option (e.g., various connector modules), which afaik would now always ignored because the execution-specific option takes precedence.
   
   I don't know whether that's a problem though. I haven't seen an increase in test instabilities when I used these commits in a WIP branch, but those also didn't ran on Azure.
   In any case we either need to remove these now-ignored forkCount settings, or adjust the poms to override it in the execution instead (or modify the properties; I think that works as well).
   I would be inclined to just remove them and see what happens.
   
   huh...there should also be some forkReuse settings that are now ignored after fee8217a65bdbcf2eed8b2c8a31852f84f1022ad. Might wanna get rid of those as well...



##########
pom.xml:
##########
@@ -95,13 +95,18 @@ under the License.
 		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 		<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 		<hadoop.version>2.8.5</hadoop.version>
+		<flink.XmxITCase>2048m</flink.XmxITCase>
+		<flink.XmxUnitTest>1024m</flink.XmxUnitTest>

Review Comment:
   I would be in favor of just hard-coding these; I don't see a need to make these configurable; you'd rather want to tweak the number of forks.



##########
pom.xml:
##########
@@ -95,13 +95,18 @@ under the License.
 		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 		<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 		<hadoop.version>2.8.5</hadoop.version>
+		<flink.XmxITCase>2048m</flink.XmxITCase>
+		<flink.XmxUnitTest>1024m</flink.XmxUnitTest>
 		<!-- Need to use a user property here because the surefire
 			 forkCount is not exposed as a property. With this we can set
 			 it on the "mvn" commandline in travis. -->
-		<flink.forkCount>1C</flink.forkCount>
+		<!-- Number of forkCounts for ITCase and UnitTest should take into account allocated memory
+			 to the jvm (-Xmx) and the available memory on the machine running the test -->
+		<flink.forkCountITCase>1C</flink.forkCountITCase>
+		<flink.forkCountUnitTest>2C</flink.forkCountUnitTest>
 		<!-- Allow overriding the fork behaviour for the expensive tests in flink-tests
 			 to avoid process kills due to container limits on TravisCI -->
-		<flink.forkCountTestPackage>${flink.forkCount}</flink.forkCountTestPackage>
+		<flink.forkCountTestPackage>${flink.forkCountITCase}</flink.forkCountTestPackage>

Review Comment:
   we should get rid of `forkCountTestPackage`; it has been identical to the usual forkCount for as long as I can remember.



##########
pom.xml:
##########
@@ -95,13 +95,18 @@ under the License.
 		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 		<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 		<hadoop.version>2.8.5</hadoop.version>
+		<flink.XmxITCase>2048m</flink.XmxITCase>
+		<flink.XmxUnitTest>1024m</flink.XmxUnitTest>
 		<!-- Need to use a user property here because the surefire
 			 forkCount is not exposed as a property. With this we can set
 			 it on the "mvn" commandline in travis. -->
-		<flink.forkCount>1C</flink.forkCount>
+		<!-- Number of forkCounts for ITCase and UnitTest should take into account allocated memory
+			 to the jvm (-Xmx) and the available memory on the machine running the test -->
+		<flink.forkCountITCase>1C</flink.forkCountITCase>
+		<flink.forkCountUnitTest>2C</flink.forkCountUnitTest>

Review Comment:
   I'm wondering if we shouldn't default to 2/4 respectively and remove the setting in the CI scripts. 1C/2C is a somewhat dangerous default; we already had contributors run into issue with that. (in fact I don't think my machine could handle that if the forks were running at full power).
   Technically not related to this PR though 🤷 



##########
flink-runtime/src/test/java/org/apache/flink/runtime/taskexecutor/TaskExecutorRecoveryITCase.java:
##########
@@ -58,7 +58,7 @@
 
 /** Recovery tests for {@link TaskExecutor}. */
 @ExtendWith(TestLoggerExtension.class)
-class TaskExecutorRecoveryTest {
+class TaskExecutorRecoveryITCase {

Review Comment:
   I would like to double-check again whether this is really necessary.
   Creating a TaskExecutor on it's own shouldn't be that expensive, and there are plenty of other unit tests doing that as well.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] pnowojski commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
pnowojski commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r850377445


##########
flink-end-to-end-tests/run-nightly-tests.sh:
##########
@@ -258,7 +258,7 @@ function run_group_2 {
     LOG4J_PROPERTIES=${END_TO_END_DIR}/../tools/ci/log4j.properties
 
     MVN_LOGGING_OPTIONS="-Dlog.dir=${DEBUG_FILES_OUTPUT_DIR} -DlogBackupDir=${DEBUG_FILES_OUTPUT_DIR} -Dlog4j.configurationFile=file://$LOG4J_PROPERTIES"
-    MVN_COMMON_OPTIONS="-Dflink.forkCount=2 -Dfast -Pskip-webui-build"
+    MVN_COMMON_OPTIONS="-Dflink.forkCountITCase=2 -Dflink.forkCountUnitTest=4 -Dfast -Pskip-webui-build"

Review Comment:
   Fixed



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r855849054


##########
flink-runtime/src/test/java/org/apache/flink/runtime/taskexecutor/TaskExecutorRecoveryITCase.java:
##########
@@ -58,7 +58,7 @@
 
 /** Recovery tests for {@link TaskExecutor}. */
 @ExtendWith(TestLoggerExtension.class)
-class TaskExecutorRecoveryTest {
+class TaskExecutorRecoveryITCase {

Review Comment:
   https://issues.apache.org/jira/browse/FLINK-27351



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on PR #19445:
URL: https://github.com/apache/flink/pull/19445#issuecomment-1110697996

   The overall benefit of this PR has diminished quite a bit with all the other cleanups being merged.
   It does look like the most of initial speedup was primarily due to enabling forkReuse and increasing the forkCount for all modules.
   
   However it still shaves of a minute or two in some modules, which is still very valuable.


-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on a diff in pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on code in PR #19445:
URL: https://github.com/apache/flink/pull/19445#discussion_r849365534


##########
pom.xml:
##########
@@ -1562,7 +1567,6 @@ under the License.
 				<configuration>
 					<!-- enables TCP/IP communication between surefire and forked JVM-->
 					<forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
-					<forkCount>${flink.forkCount}</forkCount>

Review Comment:
   I'm looking into preparing a cleanup PR that removes reuseForks, other custom forkCounts etc.
   Then we should also get a clearer picture of how much this chain specifically gets us.



-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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


[GitHub] [flink] zentol commented on pull request #19445: [FLINK-27209][build] Half the memory and double forkCount for running unit tests

Posted by GitBox <gi...@apache.org>.
zentol commented on PR #19445:
URL: https://github.com/apache/flink/pull/19445#issuecomment-1109829956

   @flinkbot run azure


-- 
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.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

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