Grafana Observability Integration

ULTIMATE

Feature Preview: This feature is available for Ultimate subscribers as a subscription add-on and may incur additional charges.

The Grafana Observability Integration streams logs and metrics from all apps into your Grafana Cloud account, or a self-hosted Grafana stack (requires a Loki & Prometheus backend), for seamless observability.

Offered as an add-on feature, this backend integration enables monitoring from external systems, faster troubleshooting, and unified visibility across your application ecosystem.

Key Capabilities

1. Platform-Level Integration

Available as an optional add-on at the platform level
Automatically captures and streams monitoring data of all deployed apps

2. Application Log Streaming

Log processing and storage require a Loki endpoint provided by your Grafana Cloud account or self-hosted Grafana stack.

Streams application logs in near real-time to the Loki endpoint of Grafana
Includes records generated at the application layer for debugging and troubleshooting, including:
- Log output of apps ranging from various levels, including errors, warnings, info and audit logs (optional).
- Run output of apps that represent the execution sequence of a flow (optional).
- Job events of apps generated by checkpoints of a flow (optional).
Enables:
- Centralized log aggregation
- Advanced search, filtering, and analytics
- Faster root cause analysis

→ Grafana Cloud log management allows teams to collect, process, and analyze large volumes of log data efficiently.

3. Application Metrics Streaming

Metric processing and storage require a Prometheus endpoint provided by your Grafana Cloud account or self-hosted Grafana stack.

Streams key runtime performance metrics to the Prometheus endpoint of Grafana, including:
- CPU utilization of app
- Memory usage of app
Provides continuous visibility into system health and performance

→ Metrics in Grafana Cloud are time-series data points used to monitor system health and performance over time

How It Works

Apps generate logs and metrics
qibb Platform automatically collects telemetry data
Data is streamed to Grafana Cloud ingestion endpoints
Grafana Cloud processes and visualizes the data in dashboards

This integration provides telemetry streaming to your Grafana stack.

Creation and management of dashboards, monitors, alerts, retention policies, and Grafana Cloud resources remain the responsibility of the customer.

Log Data Fields

Below are the most common data fields included in the records. Most fields are optional. Data fields may vary by app or platform version.

Category	Field	Description	Example
Time	timestamp	Timestamp of the record.	`1779111812501`
Log type	type	If empty: Log record. If “`run`”: Run record If “`job_event`”: Event record for a job.	-
Log Information	level	Log level, e.g. `info`, `debug`, `audit`, `warn`, `error`	`info`
	stream	Log output stream, e.g. `stdout` or `stderr`	`stdout`
	message	The log message	`The next execution time of Auto Retry task is scheduled for **********.`
App Information	qibb_appId	Id of the app.	`xmtwhy*******`
	container	The component of the app. Either `flow-app-container` (primary component that powers the low-code workflow engine which runs flow) or `flow-app-sidecar` (secondary component whic is responsible Management API processing and background tasks for jobs).	`flow-app-container`
	qibb_spaceId	Id of the space which contains this app.	`of1****`
Flow Information	flow_id	Id of the flow tab.	`abc*************`
Node Information	node_id	If of the node.	`xy5df3**********`
	node_type	Node type	`qibb-checkpoint`
	node_name	Node name	`Checkpoint`
Flow Message Information (msg)	msg_id	Id of the processed msg object.	`71ccf2bc59******`
Run Information	run_id	Id of the run.	`yy77WVc1mNJFbu*******`
Job Information	job_id	If of the job.	`a3rL4ptjpDCVj0v******`
	msg	Only applicable if type=job_event. A JSON containing event metadata. Typically includes: job_id, event, checkpoint_id, checkpoint_name, checkpoint_type, summary_plain_text	`{ "checkpoint_name":"Wait for approval", "queue_type":"WAIT", "job_id":"eh74L84596Vt3eG***", "checkpoint_id":"ae5df382911***", "checkpoint_type":"WAIT", "event_level":"INFO", "summary_plain_text":"Wait for approval: Job awaits approval.", "type":"job_event", "event":"WAIT", "attempt":0, "timestamp":"2026-05-18T13:43:32.501Z" }`
Infrastructure	qibb_cluster_id	Id of the cluster which is hosting this app.	`clx**********`

Metrics

The following metrics is the standard set which can be provided to monitor apps.

Metric names may change in future releases. Exact names vary by app and platform version.

Category	Metric	Availability	Description
Resource usage	`container_cpu_usage_seconds_total`	Default	Total CPU time consumed by the app container. Can be used to monitor CPU utilization over time.
Resource usage	`container_memory_working_set_bytes`	Default	Current memory working set of the app container. Can be used to monitor memory consumption and detect memory pressure.
Resource limits	`kube_pod_container_resource_limits{resource="cpu"}`	Optional	Configured CPU limit of the app container. Helps compare CPU usage against the configured app limit.
Resource limits	`kube_pod_container_resource_limits{resource="memory"}`	Optional	Configured memory limit of the app container. Helps compare memory usage against the configured app limit.
Lifecycle	`kube_pod_container_status_restarts_total`	Optional	Number of app container restarts. Indicates unstable apps, crash loops, or repeated failures. Note: Users may also restart containers during administrative actions like app restarts or upgrades.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_completed`	Optional	Total number of completed requests routed to the app through the qibb API gateway.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_2xx`	Optional	Number of successful responses returned by the app.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_4xx`	Optional	Number of client error responses, for example invalid requests, unauthorized requests, or missing resources.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_5xx`	Optional	Number of server error responses, typically indicating application or upstream failures.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_3xx`	Optional	Number of redirect responses returned by the app.
API / Network usage	`envoy_cluster_<app>_external_upstream_rq_time`	Optional	Request processing duration for requests routed to the app. Can be used to monitor latency percentiles where available.

Metrics are available through a curated allowlist managed by qibb’s DevSecOps team for SaaS deployments. Additional metrics can be exposed on request for PaaS deployments, depending on availability.

Labels provide customer-safe attribution and enable filtering or aggregation by app, space, and cluster level, as well as breakdown of the app components (pod or container).
Optional metrics may increase Grafana Cloud custom metric series and usage.