I wanted to step up our monitoring game by adding more helpful dashboards and noticed some shortcomings of the labels we currently use for logs and metrics. Did so with the help of Claude via Browser chat, so #llm_enabled
was still able to build this beautiful dashboard (with claude assistance..)
but it relays on regex based label replacements of this kind:
count by (image_name, image_tag) (
group by (image_name, image_tag, container_label_com_docker_stack_namespace) (
label_replace(
label_replace(container_last_seen, "image_name", "$1", "image", "([^@]+):[^:@]+.*"),
"image_tag", "$1", "image", "[^@]+:([^:@]+).*"
)
)
)
the changes introduced with prometheus.relabel "container_meta" address this issue and should make working with image tags and names more straightforward.
the second change affects the journald kernel logs to e.g. distinguish source, priority etc.:
for this I also dropped the NetworkDB stats as they produce a lot of log volume while not being very helpful at all, especially on single node deployments.
I'm currently also working on a way to add labels for version (of the recipe) and the actual recipe via abra - along with a method to set custom labels via .env
I wanted to step up our monitoring game by adding more helpful dashboards and noticed some shortcomings of the labels we currently use for logs and metrics. Did so with the help of Claude via Browser chat, so #llm_enabled
was still able to build this beautiful dashboard (with claude assistance..)

but it relays on regex based label replacements of this kind:
```
count by (image_name, image_tag) (
group by (image_name, image_tag, container_label_com_docker_stack_namespace) (
label_replace(
label_replace(container_last_seen, "image_name", "$1", "image", "([^@]+):[^:@]+.*"),
"image_tag", "$1", "image", "[^@]+:([^:@]+).*"
)
)
)
```
the changes introduced with `prometheus.relabel "container_meta"` address this issue and should make working with image tags and names more straightforward.
the second change affects the journald kernel logs to e.g. distinguish source, priority etc.:

for this I also dropped the NetworkDB stats as they produce a lot of log volume while not being very helpful at all, especially on single node deployments.
I'm currently also working on a way to add labels for `version` (of the recipe) and the actual `recipe` via abra - along with a method to set custom labels via .env
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
I wanted to step up our monitoring game by adding more helpful dashboards and noticed some shortcomings of the labels we currently use for logs and metrics. Did so with the help of Claude via Browser chat, so #llm_enabled
was still able to build this beautiful dashboard (with claude assistance..)

but it relays on regex based label replacements of this kind:
the changes introduced with
prometheus.relabel "container_meta"address this issue and should make working with image tags and names more straightforward.the second change affects the journald kernel logs to e.g. distinguish source, priority etc.:

for this I also dropped the NetworkDB stats as they produce a lot of log volume while not being very helpful at all, especially on single node deployments.
I'm currently also working on a way to add labels for
version(of the recipe) and the actualrecipevia abra - along with a method to set custom labels via .env