bug: `getting started` docs for victoria-logs-* Helm charts contain incorrect and misleading information

[valyala]

Chart name and version This concerns docs for all the VictoriaLogs-related Helm charts:

• victoria-logs-single
• victoria-logs-cluster
• victoria-logs-collector
• victoria-logs-multilevel

Describe the bug

The docs for victoria-logs-* helm charts contain getting started guides on how to quickly configure and deploy the corresponding chart with the minimum settings. Some of these guides are misleading/incorrect and/or do not contain the needed information, which may prevent from using these charts by new users. Examples:

• The docs for victoria-logs-single do not mention endpoints for logs' ingestion and querying. They also do not refer to the corresponding docs - how to send logs to VictoriaLogs and how to query logs. So, users may struggle in figuring out how to send and query logs from the just installed VictoriaLogs instance.
• The docs for victoria-logs-cluster do not mention how to send logs to the installed cluster.
• The docs for victoria-logs-collector miss the information on how to send logs to VictoriaLogs instances deployed via victoria-logs-single and victoria-logs-cluster Helm charts. This prevents from seamless setup of victoria-logs-collector.
• The docs for victoria-logs-multilevel miss working examples on how to setup global querying view over VictoriaLogs instances deployed via victoria-logs-single and victoria-logs-cluster.

Please follow all these docs as a user who doesn't know the internals of these charts, fix misleading docs and add missing information to the docs, so new users could quickly setup VictoriaLogs in Kubernetes by following these guides.

Please verify that all the commands and configs mentioned in the guides work as intended. If something doesn't work, then fix it.

[vrutkovs]

Fixing this in https://github.com/VictoriaMetrics/helm-charts/pull/2613

The docs for victoria-logs-single do not mention endpoints for logs' ingestion and querying.

This was fixes afaik - we mention query/ingestion and link to docs:

The VictoriaLogs write api can be accessed via port 9428 on the following DNS name from within your cluster:
    vl-test-victoria-logs-single-server-0.vl-test-victoria-logs-single-server.vl.svc.cluster.local.

Logs Ingestion:
  Get the Victoria Logs service URL by running these commands in the same shell:
    export POD_NAME=$(kubectl get pods --namespace vl -l "app=server" -o jsonpath="{.items[0].metadata.name}")
    kubectl --namespace vl port-forward $POD_NAME 9428

  Write URL inside the kubernetes cluster:
    http://vl-test-victoria-logs-single-server.vl.svc.cluster.local.:9428<protocol-specific-write-endpoint>

  All supported write endpoints can be found at https://docs.victoriametrics.com/victorialogs/data-ingestion/

Read Data:
  The following URL can be used to query data:
    http://vl-test-victoria-logs-single-server.vl.svc.cluster.local.:9428

However, its best to port-forward the service, as its usually backed by multi pods, survives pod restart and doesn't require additional command to retrieve the pod name. WDYT?

UPD: oops, this is for helm chart output, not the docs

[vrutkovs]

As an example for the multilevel logs I chose three single instances with different retention rate - all fed from the same source. Can you think of a better, more user-visible way to showcase it? Maybe split debug/info/errors or send different k8s component logs to different clusters?