prometheus:
tofs:
source_files:
{% raw %}prometheus-archive-install-{{ name }}-managed-service:{% endraw %}
- node_exporter.service.jinja
should result in state managing node_exporter service to look for salt://prometheus/files/default/node_exporter.service.jinja, yet it didn't. I saw in logs this:
Your setup
Formula commit hash / release tag
d9c93860385303ae89025431da7a83d48c5a6adf
Versions reports (master & minion)
Pillar / config used
where the most relevant part is
Bug details
Describe the bug
According to all my reading, and most specifically https://github.com/saltstack-formulas/prometheus-formula/blob/master/prometheus/archive/install.sls#L61-L62 pillar like
should result in state managing
node_exporter
service to look forsalt://prometheus/files/default/node_exporter.service.jinja
, yet it didn't. I saw in logs this:and the presence multiple times of
{{ name }}
was suspicious. So I made an experiment, and set pillar like so:resulting in
Suddenly all the states managing services started looking for
salt://prometheus/files/default/node_exporter.service.jinja
Steps to reproduce the bug
See above.
Expected behaviour
{{ name }}
getting properly expanded to current service name.Attempts to fix the bug
None yet, I don't know enough about jinja templating and where it applies and where it doesn't.
Additional context
Found while trying to figure out https://github.com/saltstack-formulas/prometheus-formula/issues/23