Skip to content

Memory leak by using the client #1236

@Yavor16

Description

@Yavor16

Hello, our service used this client for a long time and we didn't have any problems. We were using Java 11 but recently we migrated to Java 17. With that our problems started. Under a lot of load our services keeps crashing due to full memory. While I was doing investigation, I ssh-ed to the service and using jcmd i saw what was going on in the memory. These are classes that are loaded and not cleaned afterwards which lead to the memory leak. The number is the count of classes that are loaded. These results were taken when the service wasn't under a lot of load. During it's peak they become even higher.

761809 io.micrometer.core.instrument.ImmutableTag
409768 io.micrometer.core.instrument.Meter$Id
202341 io.netty.util.internal.shaded.org.jctools.queues.MpscArrayQueue
86861 io.micrometer.core.instrument.distribution.DistributionStatisticConfig
86857 reactor.netty.channel.MeterKey
86958. [Lio.micrometer.core.instrument.Tag;
52797 io.micrometer.core.instrument.composite.CompositeTimer
86958 io.micrometer.core.instrument.Tags

We are using 5.12.2.RELEASE. We tried also 5.12.1.RELEASE but nothing changed. I also tried to change the reactor-netty version. I tried with 1.1.21 and also 1.0.*. The result is still the same. Also tried swithing the mecrometer version but the result is the same.

Here is a link for our cf client wrapper that we are using. Here are some calls that we are using: link

Also here are classes that we are using to fetch tokens: link

Do you know where the problem could be. Thanks in advance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions