[BUG] No debug messages are logged

[macrokernel]

Describe the bug I see no debug messages in Reloader pod log:

$ kubectl -n reloader logs -f reloader-reloader-685d49466b-wl7gs
time="2025-07-18T08:29:33Z" level=info msg="Environment: Kubernetes"
time="2025-07-18T08:29:33Z" level=info msg="Starting Reloader"
time="2025-07-18T08:29:33Z" level=warning msg="KUBERNETES_NAMESPACE is unset, will detect changes in all namespaces."
time="2025-07-18T08:29:33Z" level=warning msg="namespace-selector is set, will only detect changes in namespaces with these labels: reloader.stakater.com/match_namespace."
time="2025-07-18T08:29:33Z" level=info msg="created controller for: configMaps"
time="2025-07-18T08:29:33Z" level=info msg="Starting Controller to watch resource type: configMaps"
time="2025-07-18T08:29:33Z" level=info msg="created controller for: secrets"
time="2025-07-18T08:29:33Z" level=info msg="Starting Controller to watch resource type: secrets"
time="2025-07-18T08:29:33Z" level=info msg="created controller for: namespaces"
time="2025-07-18T08:29:33Z" level=info msg="Starting Controller to watch resource type: namespaces"
time="2025-07-18T08:29:33Z" level=info msg="added namespace to be watched: opensearch"
time="2025-07-18T09:03:25Z" level=info msg="Changes detected in 'opensearch-dashboards-cert' of type 'SECRET' in namespace 'opensearch'; updated 'opensearch-dashboards' of type 'Deployment' in namespace 'opensearch'"
time="2025-07-18T09:03:29Z" level=info msg="Changes detected in 'opensearch-transport-cert' of type 'SECRET' in namespace 'opensearch'; updated 'opensearch-nodes' of type 'StatefulSet' in namespace 'opensearch'"

despite of logLevel being set to debug in Reloader deployment:

  Containers:
   reloader-reloader:
    Image:      ghcr.io/stakater/reloader:v1.4.5
    Port:       9090/TCP
    Host Port:  0/TCP
    Args:
      --log-level=debug
      --namespace-selector=reloader.stakater.com/match_namespace
      --reload-on-create=true
      --reload-strategy=annotations

To Reproduce Reloader helm values:

reloader:
  logLevel: debug
  namespaceSelector: reloader.stakater.com/match_namespace
  readOnlyRootFileSystem: true
  reloadOnCreate: true
  reloadStrategy: annotations
  watchGlobally: true

Expected behavior More details in Reloader pod logs. Perhaps Reloader simply does not log any debug messages?

Environment

• Reloader Helm chart Version: 2.1.5
• Kubernetes Version: v1.32.3+k3s1

[Evalle]

Hi, I've faced the same issue, e.g. no debug messages with reloader v 1.10 and kubernetes v1.31.6:

time="2025-07-21T13:54:25Z" level=info msg="Environment: Kubernetes"
time="2025-07-21T13:54:25Z" level=info msg="Starting Reloader"
time="2025-07-21T13:54:25Z" level=warning msg="KUBERNETES_NAMESPACE is unset, will detect changes in all namespaces."
time="2025-07-21T13:54:25Z" level=info msg="created controller for: configMaps"
time="2025-07-21T13:54:25Z" level=info msg="Starting Controller to watch resource type: configMaps"
time="2025-07-21T13:54:26Z" level=info msg="created controller for: secrets"
time="2025-07-21T13:54:26Z" level=info msg="Starting Controller to watch resource type: secrets"

[rasheedamir]

@macrokernel what information would you like to see on the debug level? And can you please tell more about the environment? Like how many namespaces? How many deployments? etc. and what issue are you trying to resolve?

[Evalle]

Hey @rasheedamir thanks for looking into it. I'm trying to understand why reloader is eating so much memory on one of my clusters. Here is some data:

Cluster	Number of Secrets	Number of Configmaps	Number of Pods	Number of Deployments	Number Of Namespaces
A	1736	1050	8944	520	299
B	1805	796	1937	550	216

On cluster A the memory consumption of reloader is around 80 MB:

On cluster B it is 16GB:

So ideally in the debug messages I would like to see what is going on there and why so much difference. Thx.

[macrokernel]

Hi @rasheedamir, I would like to see the actual restart strategy (update env, update annotations, restart) used in each update.

[rasheedamir]

@Evalle thanks for sharing details; is it same version of Reloader on both clusters and in general both clusters have similar specs?

@macrokernel Reloader uses the restart strategy you have setup on the workload yourself; if you can explain bit more with some examples what you would like to achieve then we can look into it

[macrokernel]

I am seeing exactly the same behavior with "restart" strategy as if Reloader was configured for "reload" strategy, therefore I would like to make sure that Reloader is actually using the right strategy. There could be a mistake in my configuration or a bug in Reloader. It would be nice if Reloader could log the applied strategy in every log message about the restarts it makes.

[Evalle]

@rasheedamir I don't want to abuse this issue so I created a separate one for the memory consumption: https://github.com/stakater/Reloader/issues/977