[jvm-packages] XGBoost corrupted the jvm memroy

[jinmfeng]

We run 14 xgboost in spark in parallel, but it throw out below error. it looks like the heap memory is coruppted. Do you know what's the issue? when we run single training model, there's no such error.

#

A fatal error has been detected by the Java Runtime Environment:

#

SIGSEGV (0xb) at pc=0x00007f1b028dab21, pid=23988, tid=0x00007f1acf826700

#

JRE version: OpenJDK Runtime Environment (Zulu 8.35.0.6-SA-linux64) (8.0_202-b04) (build 1.8.0_202-b04)

Java VM: OpenJDK 64-Bit Server VM (25.202-b04 mixed mode linux-amd64 compressed oops)

Problematic frame:

J 8901 C2 java.util.WeakHashMap.get(Ljava/lang/Object;)Ljava/lang/Object; (77 bytes) @ 0x00007f1b028dab21 [0x00007f1b028daa40+0xe1]

#

Core dump written. Default location: /hadoop/1/yarn/local/usercache/jinmfeng/appcache/application_1663728370843_647527/container_e3798_1663728370843_647527_01_000053/core or core.23988

#

An error report file with more information is saved as:

/hadoop/1/yarn/local/usercache/jinmfeng/appcache/application_1663728370843_647527/container_e3798_1663728370843_647527_01_000053/hs_err_pid23988.log

#

If you would like to submit a bug report, please visit:

http://www.azulsystems.com/support/

#

[LarrySunmk]

have you solved this issue?

[wbo4958]

Hi @LarrySunmk, have you encountered the same issue? if yes, please comment it with details. Recently, xgboost jvm has been refactored, please use the latest xgboost snapshot if possible from https://s3-us-west-2.amazonaws.com/xgboost-maven-repo/list.html?prefix=snapshot/ml/dmlc/xgboost4j-spark_2.12/2.2.0-SNAPSHOT/

[LarrySunmk]

您的邮件我已收到！ This is an automatic reply, confirming that your e-mail was received.Thank you