-
Notifications
You must be signed in to change notification settings - Fork 51
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Kafka 0.11 on debian:stretch #5
Conversation
Kafka docs are quite light on JVM requirements. Based on "use the latest released version of JDK 1.8" and confluentinc/cp-docker-images#21 I conclude they assume vendor is Oracle. I don't know anything about Zulu but let's do like Confluent (99f4d45). Still deviating from Confluent's image builds in many aspects. For example they are:
|
Ouch... 99f4d45 added almost 200MB to image size. 231MB instead of 37MB for the installation layer. |
as even without additional jars it needs properties files.
No longer start any service by default
RC build: TODO document
|
It would save some pull time at minor kafka upgrades if Zulu and Kafka were installed in separate layers. However due to the lack of build cache with Docker Hub automated build, the way to do this would be to maintain a Zulu image. For now let's stick with the single build strategy. |
Split docker build into JRE and Kafka
…nstall nor runtime.
Merging this without a decision on #10. Let's keep Kafka as vanilla as possible for now (though with 2MB site-docs removed). At |
Built from master: |
The docs TODOs above were fixed in master, 4a7b8da. |
Replaces #3 and #4.
At Yolean we recently evaluated Alpine (3.6) across our services. The small base image size and impressively fast
apk
saves time in development, but we had to conclude we get more glitches in production (e.g. kubernetes DNS name resolution, default shell not forwarding signals) and a bit more maintenance of our dockerfiles (e.g. 3rd party build scripts not tested with Alpine).Before I continue with the tuning in Yolean/kubernetes-kafka#30 I wanted to minimize the gambling. To do that I've decided to look at https://github.com/confluentinc/cp-docker-images as a reference setup. We still want to maintain our own Kafka image though, compatible with Docker Hub's automated build feature.