Unable to Create many threads in OCP Containers
Recently we were getting the following exception in one of our Containers which was running a Java application in Openshift 4.2. This container used to run perfectly on Openshift 3.11.
Exception in thread "main" java.lang.OutOfMemoryError: unable to create new native thread
at java.lang.Thread.start0(Native Method)
at java.lang.Thread.start(Thread.java:717
Within the containers, the ulimits looks perfectly fine. See the below image.
In Openshift 3.11 the same container could create 4096 Threads. But in Openshift 4.2 it just can create 1024 threads. Please see the Below Images,
OCP 3.11
OCP 4.2
From the above ulimits it's evident that docker agent level configurations are done. Also, I have allocated enough memory for OS to create native threads. But I have no clue where this limit is set. How can I increase this global limit? Thanks in advance.
By default, OpenShift 3 uses docker as container runtime whereas OpenShift 4 uses cri-o as container runtime.
According to Default pids_limit too low - by default there has been a limit to 1024 threads when using cri-o.
- Argocd application resource stuck at deletion
- Trying to create a namespace in an AWS EKS cluster with kubectl - Getting: Error from server (Forbidden): namespaces is forbidden
- Invalid resource manager ID in primary checkpoint record
- Don't have permission to view all the namespaces in the cluster
- Minikube single node cluster architecture
- Kubectl error: memcache.go:265] couldn’t get current server API group list: Get
- why does controller-runtime say resource 'not found' while updating resource that exists?
- Tailing few lines from huge logs of kubectl logs -f
- “kubectl exec” results in "error: unable to upgrade connection: pod does not exist"
- Ability to add exception to RKE2 securitypolicy for root pods or specific namespace
- why kubectl run onfail1 --image=alpine --restart=OnFailure -- exit 0， pod status=CrashLoopBackOff， i think should be completed
- Kubernetes - display current pods vs. capacity with kubectl
- Can Kubernetes secrets store newlines?
- Docker desktop - kubernetes failed to start
- Cannot access container from NodePort using Kubernetes ingress istio
- kubectl not working on my windows 10 machine
- How can I fix Edit cancelled, no changes made in shell
- increase proxy_send_timeout and proxy_read_timeout ingress nginx
- Kubernetes memory limit : Containers with and without memory limit on same pod
- Is there a way to specify which values file to use in a helm chart
- Private Registry for Action Pods in Openwhisk deployed on Kubernetes through helm 3
- Kubernetes stateful sets storage in AWS EBM with Helm
- Helm3 Using Lookup Function to Load a Variable
- Execute bash command in pod with kubectl?
- How do I get the External IP of a Kubernetes service as a raw value?
- Terminate istio-proxy after cronjob completion
- Allocated Resources percentage in "kubectl top node"
- How to solve helm error: "did not find expected '-' indicator"?
- kubectl command to patch a specific attribute in kubernetes storage class
- pega-web-tomcat container not starting with | Problem during method invocation (doPostInner)