Add a Source to enable telemetry collection for specific Kuberenetes workloads, or a whole namespace.
The easiest way to add a source to Odigos is to use the Odigos UI.
After installing Odigos CLI, run the following command to start the Odigos UI:
And then, in the setup page or the sources page, select the workloads / namespaces for which you want telemetry collection to be enabled.
To include all the workloads in a namespace as sources, and automatically add future-new workloads, enable the Future Apps
toggle in the UI.
Otherwise, you can enable the Select All
toggle to include all the workloads in the namespace as sources, excluding future-new workloads.
See the CLI documentation on Sources for detailed command documentation for manipulating Sources.
If you prefer to use kubectl or GitOps to manage your Kubernetes workloads, you can add a source to Odigos by creating a Source object.
Only one Source object may exist for a given workload, except in namespace instrumentation.
The following Source object enables instrumentation for a Deployment named frontend
in the default
namespace and belong to Datastream example
.
Note: Sources that do not have a Datastream label will be automatically attached to the default Datastream.
The following Source object enables instrumentation for all the workloads in the default
namespace:
Namespace Sources must have the following fields:
spec.workload.name == spec.workload.namespace
spec.workload.kind: Namespace
You can have a namespace Source and a workload Source that both enable instrumentation for a workload. In this case, if one is deleted, the other will persist instrumentation for the workload.
To exclude specific workloads from namespace instrumentation, set disableInstrumentation: true
on the workload Source. For example:
In this example, all workloads in the default
namespace will be instrumented except for the frontend
Deployment. If example-source
is deleted (or disableInstrumentation
is set to false
), the frontend
Deployment will be instrumented.
Add a Source to enable telemetry collection for specific Kuberenetes workloads, or a whole namespace.
The easiest way to add a source to Odigos is to use the Odigos UI.
After installing Odigos CLI, run the following command to start the Odigos UI:
And then, in the setup page or the sources page, select the workloads / namespaces for which you want telemetry collection to be enabled.
To include all the workloads in a namespace as sources, and automatically add future-new workloads, enable the Future Apps
toggle in the UI.
Otherwise, you can enable the Select All
toggle to include all the workloads in the namespace as sources, excluding future-new workloads.
See the CLI documentation on Sources for detailed command documentation for manipulating Sources.
If you prefer to use kubectl or GitOps to manage your Kubernetes workloads, you can add a source to Odigos by creating a Source object.
Only one Source object may exist for a given workload, except in namespace instrumentation.
The following Source object enables instrumentation for a Deployment named frontend
in the default
namespace and belong to Datastream example
.
Note: Sources that do not have a Datastream label will be automatically attached to the default Datastream.
The following Source object enables instrumentation for all the workloads in the default
namespace:
Namespace Sources must have the following fields:
spec.workload.name == spec.workload.namespace
spec.workload.kind: Namespace
You can have a namespace Source and a workload Source that both enable instrumentation for a workload. In this case, if one is deleted, the other will persist instrumentation for the workload.
To exclude specific workloads from namespace instrumentation, set disableInstrumentation: true
on the workload Source. For example:
In this example, all workloads in the default
namespace will be instrumented except for the frontend
Deployment. If example-source
is deleted (or disableInstrumentation
is set to false
), the frontend
Deployment will be instrumented.