Now report is constructed by adding more data from several stages and represents internal structure rather what API expects. + Made all function that don't interact with phabricator free standing + Now links list all files in results directory + Status icons Example comment: https://reviews.llvm.org/D71197#1779176 (links don't work).
4.8 KiB
Playbooks
deployment to a clean infrastructure
General remarks:
- GCP does not route any traffic to your services unless the service is "healthy". It might take a few minutes after startup before the services is classified as healthy. Until then you will only see some generic error message.
These are the steps to set up the build server on a clean infrastructure:
- Configure the tools on your local machine:
./local_setup.sh
- Delete the old cluster, if it still exists:
cd kubernetes/cluster ./cluster_delete.sh
- Create the cluster:
cd kubernetes/cluster ./cluster_create.sh
- Create the disk storage, if it does not yet exist:
cd kubernetes/cluster ./disk_create.sh
- SSH into the VM instance mounting the volume, find the mount point and then set
# go to the mount point of the volume cd /var/lib/kubelet/plugins/kubernetes.io/gce-pd/mounts/jenkins-home # change the permissions sudo chmod a+rwx
- Push the docker images to gcr.io:
cd containers #for each subfolder: ./build_deploy.sh <foldername>
- Deploy the stack:
cd kubernetes ./deploy.sh
- Configure it
creating basic authentication for reverse proxy
- create auth file, based on ingress-nginx documentation
cd kubernetes/reverse-proxy htpasswd -c auth <username> # enter password at prompt # add more users as required kubectl create secret generic proxy-auth --from-file=auth --namespace=jenkins
Creating docker containers on Windows
If you want to build/update/test docker container for Windows, you need to do this on a Windows machine. Here are the instructions to set up such a machine on GCP.
General hints:
- Use a GCP instance with "presistent SSD". This is much faster the the "persistent Disk".
- Try to avoid paths with white spaces in them.
- You need to configure Internet Explorer to allow downloads.
- Install Chrome, as Internet Explorer is a bit outdated.
- Install a nice IDE to edit Dockerfiles and scripts. VS Code is a good option.
- Pick a GCP Windows image with Desktop Support.
- Use the "Server Manager" application to install the "features":
- Containers
- HyperV
- Install git with the default options: https://git-scm.com/download/win
- git clone https://github.com/google/llvm-premerge-checks.git
- Register your ssh keys on the windows machine on github if you intend to push changes.
- Install docker: https://hub.docker.com/editions/community/docker-ce-desktop-windows
- You will need a DockerHub account to download the installer.
- Select "use Windows containers" during installation.
- Start the "Docker Desktop" application, it will set up the required services for you.
- Install kubectl.
- Install gcloud and set it up according to the instructions.
- Then run
gcloud auth configure-docker
to authorize docker to push images.
Check your installation by running "docker build ." in the containers/agent_windows
folder.
To push push a new container run in containers
:
powershell .\build_deploy.ps1 <container-folder>
Spawning a new windows agent
To spawn a new windows agent:
- Go to the GCP page and pick a new number for the agent.
- Update the machine name in
kubernetes/windows_agent_create.sh
. - Run
kubernetes/windows_agent_create.sh
- Go to the GCP page again
- login to the new machine via RDP (you probably need to set the i).
- In the RDP session: run these commands in the CMD window to start the docker container:
powershell
Invoke-WebRequest -uri 'https://raw.githubusercontent.com/google/llvm-premerge-checks/master/kubernetes/windows_agent_bootstrap.ps1' -OutFile windows_agent_bootstrap.ps1
.\windows_agent_bootstrap.ps1
Testing scripts locally
Build and run agent docker image sudo build_run.sh agent-debian-testing-clang8-ssd /bin/bash
.
Within a container set environment variables similar to pipeline.
Additionally set WORKSPACE
, PHID
and DIFF_ID
parameters. Set CONDUIT_TOKEN
with your personal one from https://reviews.llvm.org/settings/user/<USERNAME>/page/apitokens/
.