FAQ
Common questions about Telflo.
General
What is Telflo? A control plane for OpenTelemetry Collector configuration. You build pipelines on a canvas, Telflo generates standard Collector YAML, and you test and deploy the result to collectors. See the Introduction.
Do I need an account? You can explore the public editor and browse flows without an account. You must sign in to save, test, publish, or deploy. See the Quickstart.
Can I start from a prebuilt configuration? Yes. Flows are prebuilt, validated collector configurations, each solving one problem. Using one creates your own independent configuration. See Flows.
Can I import an existing Collector config? Yes. Paste it into the YAML view of a new configuration, and Telflo builds the canvas from it.
Does Telflo have an API? Yes. Organization admins create API tokens in Settings > API Tokens, and clients use them as bearer tokens to list and read the organization's configurations and validate collector YAML. See API tokens.
Pipelines
Which components are available? Receivers, processors, exporters, connectors, and extensions. The available set depends on the collector version you target. See Components.
How are pipelines defined? By the edges you draw. Each edge belongs to one pipeline (traces, metrics, logs, or a custom name), and a node can belong to several pipelines. See Canvas & pipelines.
What is the difference between validation and testing? Validation checks that a configuration is well-formed and schema-valid. Testing runs sample telemetry through it on a real collector and checks the output. See Testing.
Versions and fleets
What is the difference between saving and publishing? Saving records a private version in the editor. Publishing marks a version as a release that can be deployed. See Versions & publishing.
How do collectors get my configuration? They pull it from Telflo. You create a fleet, install collectors from its deploy dialog, and they connect and apply the assigned configuration. See Fleet management.
How do I roll back? On the fleet's Deploy Groups tab, edit the deploy group and assign an earlier published version. Collectors converge on it. See Update the configuration.
Can different collectors in one fleet run different configurations? Yes. A server fleet's collectors are organized into deploy groups, and each group has its own assigned configuration version and variable group. A single instance can also be pinned to a specific version. See Create a fleet.
Which platforms can I install collectors on? Collectors run on Linux, on arm64 or amd64 (x86_64), as a container or directly on a host, and in Kubernetes clusters through a Kubernetes fleet. See Install collectors.
Can Telflo manage collectors in a Kubernetes cluster? Yes. A Kubernetes bridge fleet installs one bridge into the cluster; the bridge connects to Telflo and manages the collectors running under the OpenTelemetry Operator, with configurations assigned per collector. A Kubernetes supervisor fleet, available in private preview, runs supervisor-managed collectors installed by Helm, without the Operator. See Kubernetes fleets.
Do I have to build a container image to run a collector in Docker? No. The deploy dialog's Container image (GHCR) option runs Telflo's prebuilt collector image with the fleet's credentials. See Install collectors.
AI assistant
Why don't I see the AI assistant? AI must be enabled in Settings > AI, and access may need to be requested first. When AI is enabled, the assistant opens from the AI button in the editor toolbar. See AI access.
Does the AI change my configuration? In Build mode it can, applying edits to the canvas for you to review. Ask mode only answers questions. See Modes.
Security
How do I keep credentials out of my configuration?
Reference them as environment variables with ${env:NAME}, so the value comes from the host environment rather than the stored YAML. For reusable non-secret values such as endpoints and hostnames, use organization variables: typing @ in a field inserts a $orgvar: token that Telflo resolves for each collector when it checks in. A user variable, inserted with #, places its value directly into the configuration.
Last updated on