Observability

You want to see what your applications are doing — logs, metrics, dashboards,
alerts. Some of that is available to every team by default in SVIEW; the rest
comes with a dedicated observability stack you can request.

By default, PushOps monitors platform health for CTO2B's own operational use,
and that monitoring is not visible to your team. A dedicated observability
stack — your own Grafana, Prometheus, and Loki — is opt-in and provisioned on
request.

What you can see without a dedicated stack

Every team has the following in SVIEW, with no additional setup:

• Application logs — open any Component and go to the Logs tab to view
  and search recent output. The in-UI viewer retrieves up to 6 hours of
  history.
• Service health — the Health check view shows cross-environment health
  status for your Components.
• Compliance reports — the Reports section shows node utilisation and
  compliance posture.
• Costs — spend broken down by namespace.
• Audit Logs — every platform action taken by your team.

Requesting your own observability stack

Raise a request in your Slack channel asking for a dedicated observability
instance. CTO2B will provision a Grafana, Prometheus, and Loki stack scoped to
your environments.

Once provisioned, your team can:

• Build and save custom Grafana dashboards for your applications
• Query logs with Loki's LogQL, beyond the 6-hour in-UI window
• Set up alert rules scoped to your services
• Access metrics for all Components running in your environments
• View ingress (Traefik) metrics in your Grafana dashboards

Getting people access to Grafana

Grafana access is granted per person and is time-limited — there is no single
shared login. Two levels are available, and you can request either one for a
given team member:

• Admin — build and edit dashboards, configure alert rules and notification
  channels, manage data sources.
• Read-only — view dashboards and explore logs and metrics, without
  changing any configuration.

Access is requested in SVIEW under Access requests → New on the
observability application, selecting the role that matches the level you need,
and then reached through Teleport once approved. Tell CTO2B who needs which
level and for how long — from a few hours up to a few months — so the request
can be approved. Grants can be renewed or extended.

If you do not have a SVIEW and Teleport account yet, or your Teleport identity
is not connected, follow the login and identity-connection steps in Access
Requests first. That is a one-time setup; afterwards every new grant only
needs the request above.

Querying logs in Grafana

With a dedicated stack, application logs are in Loki:

1. Go to Explore in the Grafana sidebar.
2. Select the Loki data source.
3. Filter by namespace and application label:

{namespace="your-namespace", app="your-app-name"}

To search within the logs:

{namespace="your-namespace"} |= "error"

Platform alerting

CTO2B's own monitoring sends alerts to your dedicated Slack channel when
platform-level issues are detected — unhealthy nodes, persistent failures,
storage pressure, and similar events. You do not need to configure this; it is
part of the default service.

If an alert fires and you are unsure whether it is a platform issue or
something in your application, raise a request and the team will help diagnose.

Related articles

• Access Requests — logging in to SVIEW, connecting Teleport, raising a request
• Grafana — Querying Google Managed Prometheus
• Audit Logs