[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user Ngone51 commented on the issue: https://github.com/apache/spark/pull/23223 Hi @tgravescs , I tried it, but found it's difficult to produce KILLED_BY_RESOURCEMANAGER exit status. I followed [YARN-73](https://issues.apache.org/jira/browse/YARN-73) [YARN-495](https://issues.apache.org/jira/browse/YARN-495), but things didn't go as I expected. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user Ngone51 commented on the issue: https://github.com/apache/spark/pull/23223 > it would be interesting to test it further to see if it does. @tgravescs Yeah, I have the same thought. I'd like to try it, but I can not guarantee that I can achieve it... Because I have never done this kind of test before. I'll try my best. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user tgravescs commented on the issue: https://github.com/apache/spark/pull/23223 if you aren't seeing actual issues with this I guess it would be interesting to test it further to see if it does. I can see spark blacklisting when it shouldn't for exit codes like you mention (KILLED_BY_RESOURCEMANAGER) . so I guess I would like to see someone test this further and determine if that happens. If it does we should change to bug and put into 2.4.1. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user SparkQA commented on the issue: https://github.com/apache/spark/pull/23223 **[Test build #99757 has finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/99757/testReport)** for PR 23223 at commit [`2d1c27a`](https://github.com/apache/spark/commit/2d1c27aa1cf94a9a4a524ddc16670a25c0c3b41d). * This patch passes all tests. * This patch merges cleanly. * This patch adds no public classes. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Test PASSed. Refer to this link for build results (access rights to CI server needed): https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/99757/ Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Merged build finished. Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Test PASSed. Refer to this link for build results (access rights to CI server needed): https://amplab.cs.berkeley.edu/jenkins//job/testing-k8s-prb-make-spark-distribution-unified/5796/ Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Merged build finished. Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user SparkQA commented on the issue: https://github.com/apache/spark/pull/23223 **[Test build #99757 has started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/99757/testReport)** for PR 23223 at commit [`2d1c27a`](https://github.com/apache/spark/commit/2d1c27aa1cf94a9a4a524ddc16670a25c0c3b41d). --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user Ngone51 commented on the issue: https://github.com/apache/spark/pull/23223 > it looks like its only going to blacklist the node for the AM, not other nodes for general containers. @squito Yarn have blacklist for AM when config `am-scheduling.node-blacklisting-enabled`=true, and have `ContainerFailureTracker` for general containers(haven't find a config for it). --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user Ngone51 commented on the issue: https://github.com/apache/spark/pull/23223 > Are you seeing actual issues with this blacklisting when it shouldn't? Unfortunately, no. @tgravescs @squito --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user Ngone51 commented on the issue: https://github.com/apache/spark/pull/23223 > I mean if node blacklisting in Spark would be perfectly aligned to YARN then it would be just redundant to have it in Spark in the first place. This change seems result in *perfectly* aligned to YARN for node blacklisting in Spark, but my original thought is that some exit status (e.g. KILLED_BY_RESOURCEMANAGER ), currently, should not lead to a node blacklisting. So, actually, *perfectly* aligned to YARN is not the real target of this change, and we can also make some custom strategy for Spark. > Take for example disk failure. For spark task level backlisting, is it should be delegated to **schedulerBlacklist** in YarnAllocatorBlacklistTracker ? And it seems ContainerExitStatus.DISKS_FAILED in YARN is not same with Spark tasks' disk failure. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user squito commented on the issue: https://github.com/apache/spark/pull/23223 @attilapiros > I mean if node blacklisting in Spark would be perfectly aligned to YARN then it would be just redundant to have it in Spark in the first place. I'm not super familiar with exactly how the blacklisting works in yarn itself -- it looks like its only going to blacklist the node for the AM, not other nodes for general containers. I don't totally follow where the `KILLED_BY_RESOURCEMANAGER` status is generated, but it does seem like a good idea to protect against this, maybe there is a race where the container is created by the RM, but before it reports back to the driver it gets killed with KILLED_BY_RESOURCEMANAGER. (another reason I'm curious if @Ngone51 is actually seeing this cause problems, or just noticed a case to improve.) --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Test PASSed. Refer to this link for build results (access rights to CI server needed): https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/99733/ Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user SparkQA commented on the issue: https://github.com/apache/spark/pull/23223 **[Test build #99733 has finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/99733/testReport)** for PR 23223 at commit [`65a70dc`](https://github.com/apache/spark/commit/65a70dcbb7993731104deab2592a5b969a31414e). * This patch passes all tests. * This patch merges cleanly. * This patch adds no public classes. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Merged build finished. Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user tgravescs commented on the issue: https://github.com/apache/spark/pull/23223 the approach here makes sense. Are you seeing actual issues with this blacklisting when it shouldn't? I could see that possible there and if so we should move this to defect and make sure it goes into 2.4.1 --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Merged build finished. Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user SparkQA commented on the issue: https://github.com/apache/spark/pull/23223 **[Test build #99733 has started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/99733/testReport)** for PR 23223 at commit [`65a70dc`](https://github.com/apache/spark/commit/65a70dcbb7993731104deab2592a5b969a31414e). --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Test PASSed. Refer to this link for build results (access rights to CI server needed): https://amplab.cs.berkeley.edu/jenkins//job/testing-k8s-prb-make-spark-distribution-unified/5773/ Test PASSed. --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user tgravescs commented on the issue: https://github.com/apache/spark/pull/23223 ok to test --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Can one of the admins verify this patch? --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] spark issue #23223: [SPARK-26269][YARN]Yarnallocator should have same blackl...
Github user AmplabJenkins commented on the issue: https://github.com/apache/spark/pull/23223 Can one of the admins verify this patch? --- - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
