eshepelyuk/cmak-docker

Too long startup time?

Closed this issue · 7 comments

when I start your docker images,It need more than 5mins to access the page;
I have a good suggestion,
change this file:
/usr/local/openjdk-11/conf/security/java.security
securerandom.source=file:/dev/random   to  securerandom.source=file:/dev/urandom,
It starts right away

Looking forward to further communication;

Seems we can just use-Dsecurerandom.source=file:/dev/urandom option in cmd. Feel free to provide PR, to add it to entrypoint.

Thanks for your help,But I didn't go well
when I run this images and add -Dsecurerandom.source to my compose file ,it's works,
but I reboot my ecs
image
I can't open the page;
docker logs cmak:
image
I don't know what to do,
it's my docker-composer.yml:

version: '3.6'
services:
  kafka_manager:
    image: hlebalbau/kafka-manager:stable
    container_name: "cmak"
    restart: always
    ports:
      - "9005:9000"
    command:
      - "-Dcmak.zkhosts=172.17.0.1:2181"
      - "-DbasicAuthentication.enabled=true"
      - "-DbasicAuthentication.username=username"
      - "-DbasicAuthentication.password=password123"
      - "-Dsecurerandom.source=file:/dev/./urandom"

@zengxc-1996 how do you deploy to ECS? Using a docker-compose file, or some other tool?

TBH, I doubt this issue is related to an old lack of entropy . I observed those delays randomly on different clusters and for me it seems like an issue in CMAK itself.

So, should we do something here, or just close?

So, should we do something here, or just close?

For my understanding - this issue can't be addressed completely in initially proposed way.
I still some time observer very slow startup sometimes , quite sporadically.
With the lack of entropy - container couldn't start att all.
So I think this is rather CMAK issue with connection handling.

okay, close for now

future reader, if you have a solution, please, reopen :)