Upbound provides a control plane framework for businesses to accelerate their journey building custom clouds on top of arbitrary cloud resources, making the process quicker, more flexible, and reducing cognitive load for users. With Upbound, users can reduce risk and make themselves audit-ready by enforcing governance and compliance with the Universal Crossplane.
Universal Crossplane users can offer their own internal cloud API to their teams through composite resource definitions. Their custom cloud service can then map that API to downstream providers. The providers create and manage the lifecycle of the external cloud resources orchestrated through Universal Crossplane when a user claims resources. Amazon AWS, Microsoft Azure, and Google GCP are out of the box supported cloud providers.
This Datadog integration facilitates the monitoring of Universal Crossplane and provider Kubernetes pods. Metrics are sent from Upbound's pods in the self hosted customer environment to the customer's Datadog account. They track the health of this self-managed infrastructure provisioning and resource lifecycle management ecosystem. The metrics inform about performance and resource consumption. This helps validate Universal Crossplane management cluster sizing and enables optimization.
This check looks for UXP and provider pods by default
in the upbound-system Kubernetes
namespace. The Upbound and provider
pods expose metrics on the /metrics URL from port 8080 in
a Prometheus compatible format.
The Datadog agent requires permissions to discover UXP and provider pods. Create a service account, a cluster role, and a cluster role binding to grant those permissions following the example below.
Note that the Datadog agent in the configuration
below resides in an arbitrary monitoring namespace.
Your agent may reside in a namespace of your choice.
Configure the service account and cluster role accordingly.
apiVersion: v1
kind: ServiceAccount
metadata:
name: datadog-upbound
namespace: monitoring
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: apiserver-cluster-role
namespace: monitoring
rules:
- apiGroups:
- ""
resources:
- pods
verbs: ["get", "list" ]
- nonResourceURLs: ["/metrics"]
verbs: ["get"]
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: apiserver-cluster-role-binding
subjects:
- namespace: monitoring
kind: ServiceAccount
name: datadog-upbound
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: apiserver-cluster-role
The Upbound UXP integration is not included in the Datadog Agent package, so you need to install it manually.
For Agent v7.21+ / v6.21+, follow the instructions below to install the check on your host.
-
Using the Dockerfile example in Use Community Integrations, build a custom image of the Agent that includes the Upbound integration.
-
Run the following command to install the Agent integration:
datadog-agent integration install -t datadog-upbound-uxp==1.0.0
- Edit the
upbound_uxp.d/conf.yamlfile, in theconf.d/folder at the root of your Agent's configuration directory to start collecting your Upbound UXP performance data.
See example configuration parameters below.
## Options defined here are available for Upbound UXP integrated instances.
#
init_config:
## @param service - string - optional
## Attach the tag `service:<SERVICE>` to every metric, event,
## and service check emitted by this integration.
##
## Additionally, this sets the default `service` for every log source.
#
# service: <SERVICE>
## Every instance is scheduled independently of the others.
#
instances:
## @param min_collection_interval - number - optional - default: 15
## This changes the collection interval of the check. For more information, see:
## https://docs.datadoghq.com/developers/write_agent_check/#collection-interval
#
#
# min_collection_interval: 15
## @param uxp_url - string - optional - default: "/metrics"
## This changes at which path metrics are scraped.
#
# uxp_url: "/metrics"
## @param uxp_port - string - optional - default: "8080"
## This changes at which port metrics are scraped.
#
# uxp_port: "8080"
## @param uxp_hosts - array of strings - optional - default: []
## Specifies hosts from which to collect metrics.
## Note, the Datadog Agent discovers Crossplane
## and provider pods automatically. A usecase is for
## out of cluster agent tests.
#
# uxp_hosts: [
# "localhost"
# ]
## @param verbose - boolean - optional - default: false
## The agent provides additonal logging output when
## verbose is selected.
#
# verbose: false
## @param namespace - string - optional - default: upbound-system
## The namespace where the Crossplane and Provider pods reside.
#
# namespace: "upbound-system"
## @params metrics_default - string - optional - default: min
## Easy selection of metrics default set. Values can be
## none # indicates no default metrics and implies that custom
## # selection will be provided in "metrics" field.
## min # indicates opinionated min set
## more # indicates min set plus additional metrics
## max # indicates all metrics that the Crossplane pod and providers
## # emit will be collected and forwarded to the Datadog
## # organization identified by the Datadog API key specified during
## # agent installation.
#
# metrics_default: "min"
## @params metrics_ignore_pod_annotations - boolean - optional - default: true
## By default the agent determines which metrics to scrape based
## on this configuration file. When metrics_ignore_pod_annotations is set
## to false, then the agent will also apply pod annotations.
#
# metrics_ignore_pod_annotations: true
## @params prefix - string - optional - default: ""
## Prefix that will be inserted between uxp. and (mapped)
## metrics name.
#
# metrics_prefix: ""
## @params metrics - array of strings - optional - default: []
## Specification of individually picked metrics.
## For a full list of metrics set metrics_default to "max" and
## look at your Datadog metrics explorer in the Datadog console.
#
# metrics: [
# {"go_goroutines": "company_prefix_go_goroutines"},
# # only one value needed when name mapping is not needed
# {"go_memstats_heap_alloc_bytes"},
# ]
See conf.yaml.example for a generic configuration example.
-
Pod Annotations You may annotate your Crossplane and Provider pods directly with a subset of the metrics you wish to collect from them.
A sample annotation may look as follows:
customAnnotations:
# Picked up by Universal Crossplane Datadog Integration
# Pattern to match: ad.datadoghq.com/uxp.<pod-name without pod hash>.instances
ad.datadoghq.com/uxp.crossplane.instances: |
[
{
"metrics": [
{"controller_runtime_active_workers": "crossplane_controller_runtime_active_workers"},
{"controller_runtime_reconcile_errors": "controller_runtime_reconcile_errors_total"},
{"controller_runtime_reconcile": "controller_runtime_reconcile_total"},
{"go_gc_duration_seconds": "crossplane_go_gc_duration_seconds"},
{"go_goroutines": "crossplane_go_goroutines"}
{"go_memstats_mcache_inuse_bytes": "crossplane_go_memstats_mcache_inuse_bytes"},
{"go_memstats_mcache_sys_bytes": "crossplane_go_memstats_mcache_sys_bytes"},
{"go_memstats_mspan_inuse_bytes": "crossplane_go_memstats_mspan_inuse_bytes"},
{"go_memstats_mspan_sys_bytes": "crossplane_go_memstats_mspan_sys_bytes"},
{"go_memstats_next_gc_bytes": "crossplane_go_memstats_next_gc_bytes"},
{"go_memstats_other_sys_bytes": "crossplane_go_memstats_other_sys_bytes"},
{"go_memstats_stack_inuse_bytes": "crossplane_go_memstats_stack_inuse_bytes"},
{"go_memstats_stack_sys_bytes": "crossplane_go_memstats_stack_sys_bytes"},
{"go_memstats_sys_bytes": "crossplane_go_memstats_sys_bytes"},
{"go_threads": "crossplane_go_threads"},
{"leader_election_master_status": "crossplane_leader_election_master_status"},
{"process_cpu_seconds": "crossplane_process_cpu_seconds_total"},
{"process_max_fds": "crossplane_process_max_fds"},
{"process_open_fds": "crossplane_process_open_fds"},
{"process_resident_memory_bytes": "crossplane_process_resident_memory_bytes"},
{"process_start_time_seconds": "crossplane_process_start_time_seconds"},
{"process_virtual_memory_bytes": "crossplane_process_virtual_memory_bytes"},
{"process_virtual_memory_max_bytes": "crossplane_process_virtual_memory_max_bytes"},
{"rest_client_requests": "crossplane_rest_client_requests_total"},
{"workqueue_adds": "crossplane_workqueue_adds_total"},
{"workqueue_depth": "crossplane_workqueue_depth"},
{"workqueue_retries": "crossplane_workqueue_retries_total"},
{"workqueue_unfinished_work_seconds": "crossplane_workqueue_unfinished_work_seconds"},
{"workqueue_work_duration_seconds": "crossplane_workqueue_work_duration_seconds_bucket"}
]
}
]
Run the Agent's status subcommand and look for upbound_uxp under the Checks section.
See metadata.csv for a list of metrics provided by this integration.
Upbound UXP does not include any service checks.
Upbound UXP does not include any events.
Need help? Contact Datadog support.