draughtsman
draughtsman copied to clipboard
An in-cluster agent that handles Helm based deployments
This PR is generated by https://github.com/giantswarm/github, it will be merged automatically once all requirements are met
This project looks super interesting, but it is difficult to understand how it all works. A sample would be great, along with a document that walks the user through it.
Draughtsmans helm gets killed because of RAM. ``` r7vme@0000af4ae0da2a3d ~ $ dmesg | grep "Killed process" [1993891.355272] Killed process 25939 (pause) total-vm:1024kB, anon-rss:4kB, file-rss:0kB, shmem-rss:0kB [2076617.488262] Killed process 30674 (pause)...
Current process with installing draughtsman is strange (described here - https://github.com/giantswarm/giantswarm/wiki/Draughtsman-Installation#install-draughtsman-chart) 1. prepare draughtsman values (config map an secret) 2. download helm package 3. untar helm package 4. replace `values.yaml`...
And respond in https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/kubernetes-sig-apps/WEX67CghRQg/a_je914_AQAJ afterwards
See https://github.com/giantswarm/giantswarm/issues/1647
When we start, and have no etag cache, we can end up requesting a large number of deployment statuses. Makes sense to limit this to the last 3 deployment events,...