fc1851cdc1
This adds a '--io-setup N' command line option, which users can pass to specify whether they want to run the "scylla_io_setup" script or not. This is useful if users want to specify I/O settings themselves in environments such as Kubernetes, where running "iotune" is problematic. Fixes #6587
Building a CentOS-based Docker image
Running
docker build -t <image-name> .
in this directory will build a Docker image of Scylla. You can then run it with, for example,
docker run --name scylla -d -p 9042:9042 -t <image name>
However, it is important to note that while the resulting image will contain some scripts taken from this directory, the actual Scylla executable will not be taken from this build directory. Instead, our Dockerfile downloads the Scylla executable and other Scylla tools (e.g., JMX, nodetool, etc.) from http://downloads.scylladb.com/. If you want to build a Docker image which includes a Scylla executable which you compiled yourself, please refer to the next section.
Docker image with a self-built executable
The following instructions will allow you to build a Docker image which contains a combination of some tools from the nightly build in http://downloads.scylladb.com/ (as described above) but with a Scylla executable which you build yourself.
The following instructions are currently messy, but we hope to one day make them as simple as "ninja docker".
Do the following in the top-level Scylla source directory:
-
Build your own Scylla in whatever build mode you prefer, e.g., dev.
-
Run
./reloc/build_reloc.sh --mode dev -
Run
./reloc/build_rpm.sh --reloc-pkg build/dev/scylla-package.tar.gz -
cd to
dist/docker/redhat -
Docker stubbornly refuses to allow using files from outside the current directory in preparing images, so we must copy the RPMs prepared above into the current directory (a symbolic link would not work):
rm -r rpms; cp -a ../../../build/redhat/RPMS/x86_64 rpms -
Add the following lines near the end of Dockerfile, after the
RUN curl:COPY rpms /rpms RUN yum install -y /rpms/*.rpm -
Finally, run
docker build -t <image-name> .