Vincent Viallet
Vincent Viallet
Similarly to Ubuntu, `/var/run` or `/run` on CentOS 7 is a `tmpfs` based file system and get emptied on reboot. Adding the creation step of that PIDDIR folder in the...
Currently only the step `result` (return code) is propagated from step to step. We need a way to capture / consume the output from step to step (eg. text, json,...
To not have to create everything from the UI, we shoudl offer the ability to users to create a new pipeline, and define a remote source to pull the pipeline...
Currently the pipelines are files on the server. We need to add the ability to create a pipeline from the UI directly. User story - creation: - can click a...
User story: - user can click and edit the existing pipeline configuration (codemirror editor or alike) - user can then save
Currently, we only know when the pipeline got run, and whether it was success/failure. There is no way other than checking in the logs what were the supplied parameters. Can...
Currently it takes the time of the server (?) - and we can't set the timezone to be offering the current time. See what to do.
UI points at wiki that lacks any user info. Unclear what is the relation between workspace / pipelines def, etc.
Nice to have feature; especially to allow having 1 central front-facing pipelines and *executioners* in different locations - sometime internal within a LAN. Proposed (high-level) implementation could be: - have...
The pipeline_context is only instantiated on start, updated via the prompt / webhook, and on completion of the execution of a task. it currently is not possible to update that...