java 11 HttpClient leads to endless SSL loop even with disabled TLSv.1.3

Question

I facing a problem with standard java http client. By some reason that I don't know yet it go into infinite loop and utilize all cpus I have. Below is an example stack trace and a method where the infinity loop is running.

"HttpClient-1-Worker-34" #232 daemon prio=5 os_prio=0 cpu=37704.71ms elapsed=231.22s tid=0x00007f1580279000 nid=0x31fc runnable  [0x00007f1575059000]
   java.lang.Thread.State: RUNNABLE
    at jdk.internal.net.http.common.SSLFlowDelegate$Writer.processData([email protected]/SSLFlowDelegate.java:758)
    at jdk.internal.net.http.common.SSLFlowDelegate$Writer$WriterDownstreamPusher.run([email protected]/SSLFlowDelegate.java:645)
    at jdk.internal.net.http.common.SequentialScheduler$CompleteRestartableTask.run([email protected]/SequentialScheduler.java:147)
    at jdk.internal.net.http.common.SequentialScheduler$SchedulableTask.run([email protected]/SequentialScheduler.java:198)
    at jdk.internal.net.http.common.SequentialScheduler.runOrSchedule([email protected]/SequentialScheduler.java:271)
    at jdk.internal.net.http.common.SequentialScheduler.runOrSchedule([email protected]/SequentialScheduler.java:224)
    at jdk.internal.net.http.common.SSLFlowDelegate$Writer.triggerWrite([email protected]/SSLFlowDelegate.java:722)
    at jdk.internal.net.http.common.SSLFlowDelegate.doHandshake([email protected]/SSLFlowDelegate.java:1024)
    at jdk.internal.net.http.common.SSLFlowDelegate.doClosure([email protected]/SSLFlowDelegate.java:1094)
    at jdk.internal.net.http.common.SSLFlowDelegate$Reader.unwrapBuffer([email protected]/SSLFlowDelegate.java:500)
    at jdk.internal.net.http.common.SSLFlowDelegate$Reader.processData([email protected]/SSLFlowDelegate.java:389)
    - locked <0x0000000613d63450> (a java.lang.Object)
    at jdk.internal.net.http.common.SSLFlowDelegate$Reader$ReaderDownstreamPusher.run([email protected]/SSLFlowDelegate.java:263)
    at jdk.internal.net.http.common.SequentialScheduler$SynchronizedRestartableTask.run([email protected]/SequentialScheduler.java:175)
    - locked <0x0000000613d63488> (a java.lang.Object)
    at jdk.internal.net.http.common.SequentialScheduler$CompleteRestartableTask.run([email protected]/SequentialScheduler.java:147)
    at jdk.internal.net.http.common.SequentialScheduler$SchedulableTask.run([email protected]/SequentialScheduler.java:198)
    at java.util.concurrent.ThreadPoolExecutor.runWorker([email protected]/ThreadPoolExecutor.java:1128)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run([email protected]/ThreadPoolExecutor.java:628)
    at java.lang.Thread.run([email protected]/Thread.java:834)

I thought that root cause of the problem is this one. But after disabling TLSv1.3 the issue is not go away.

I checked if TLSv1.3 is disable in the following way using debug at start up of application:

Here is JVM parameters I pass in:

-Djdk.tls.disabledAlgorithms="TLSv1.3"
-Dhttps.protocols="TLSv1,TLSv1.1,TLSv1.2"
-Djdk.tls.client.protocols="TLSv1,TLSv1.1,TLSv1.2"
-Djdk.tls.server.protocols="TLSv1,TLSv1.1,TLSv1.2"

to be sure that TLSv1.3 is disabled.

Please, any ideas what can be a reason of the issue?

Anatolian

Is it host specific or does it occur on various hosts? – Pampero

It occurs at least on both localhost and stg hosts – Anatolian

@Anatolian did you resolved it? – Kolkhoz

Sorry for been so late. Yeah, I spend a lot of time trying to solve the issue but unfortunately the only solution that worked for me was migration to JDK12. Hope I helped you. – Anatolian

I believe I ran into the same issue with jdk 11.0.2. I tried to upgrade to 11.0.6 but it still happened. Upgrading to 13.0.2 seems to solve the issue though. Upgrading to solve such a problems isn't necessarily weird, but I would have guessed, that such a big problem will be backported. So maybe I did something wrong. Anyway 13 seems to work – Nonsuit

@SimonLenz the fix was backported to 11.0.8 bugs.openjdk.java.net/browse/JDK-8241054https://… – Polymerize