Hello Team,

I am executing DistCP commands in my SpringBoot application to copy files
to AWS S3 buckets.
JAR used for this integration -

   -  hadoop-aws-2.10.1.jar
   -  aws-java-sdk-bundle-1.11.837.jar

My application goes out of memory and I have to enforce GC to clear all
memory. My investigation says most of the hadoop classes are retained by
AWS SDK classes even till the last leg of distcp execution. Apparently, GC
can not garbage collect classes until they are released by AWS classes.
Even static variables are also referenced by AWS SDK classes. It is leading
to memory leak and as I run my application in thread I see many more
instances are getting retained causing memory leak further.

I am stuck trying to resolve these memory leaks. Any help on this would be
appreciated.

Thank You,
Shashank

Reply via email to