This project derives from work by the Kubernetes team at Wayfair of general interest to users of Tremor and Kubernetes. The original work was by Njegos Railic from Wayfair's Kubernetes SRE Team.
Support for OpenShift was investigated by Anton Whalley.
The project can be used as a template for deploying standard tremor installation for Kubernetes. This project can be forked and modified for specific use cases.
Generally useful enhancements, improvements, fixes and changes that are encouraged. If you have made a change you would like to contribute please submit a pull request!
This project assumes familiarity with Tremor, Kubernetes and Helm. Additionally, it
assumes that you already have helm and kubectl binary installed.
For now we don't have a public Helm registry configured and you would have to download helm chart and add it to your Helm registry.
You would have to manually create helm package:
$ helm package charts/tremor
Successfully packaged chart and saved it to: /Users/njegosrailictremor-k8s-helm/tremor-0.1.0.tgz
helm repo updateSee helm repo for command documentation.
We may need to install tremor directly from a package and recreate the existing release. After the package is created we can execute the following commands:
$ helm install [RELEASE_NAME] tremor-kuberenetes-0.1.0.tgz
$ helm delete [RELEASE_NAME]Please check the following command for Helm v3:
# Helm install and create namespace
$ helm install -n tremor [RELEASE_NAME] [HELM_REPO_NAME]/tremor --create-namespace
# Helm install with tremor namespace already created
$ helm install -n tremor [RELEASE_NAME] [HELM_REPO_NAME]/tremor
See parameters below.
See helm install for command documentation.
Check that our release has deployed ok:
$ kubectl get pod -n tremor-test
NAME READY STATUS RESTARTS AGE
tremor-test-84f659495d-5v759 1/1 Running 0 75mInspect the log output for any other deployment errors:
$ kubectl logs tremor-test-84f659495d-5v759 -n tremor-test
+ [ -n ]
+ [ -z ]
+ TREMOR_DIR=/etc/tremor
...
+ ARGS=server run --logger-config /etc/tremor/logger.yaml -f /etc/tremor/config/*.yaml /etc/tremor/config/*.trickle
+ exec /usr/bin/tini /tremor -- server run --logger-config /etc/tremor/logger.yaml -f /etc/tremor/config/config.yaml /etc/tremor/config/main.trickle
tremor version: 0.11.4 HEAD:0c643f419124c0484a817202e0d7a79a33996dd3
tremor instance: tremor
rd_kafka version: 0x000000ff, 1.5.0
allocator: snmalloc
2021-06-03T11:24:44.944326589+00:00 INFO tremor_runtime::version - tremor version: 0.11.4 HEAD:0c643f419124c0484a817202e0d7a79a33996dd3
2021-06-03T11:24:44.944565360+00:00 INFO tremor_runtime::version - rd_kafka version: 0x000000ff, 1.5.0
2021-06-03T11:24:44.945046744+00:00 INFO tremor_runtime::onramp - Onramp manager started
2021-06-03T11:24:44.945343680+00:00 INFO tremor_runtime::offramp - Offramp manager started
2021-06-03T11:24:44.945491373+00:00 INFO tremor_runtime::pipeline - Pipeline manager started
2021-06-03T11:24:45.035620164+00:00 INFO tremor_runtime::system - Binding pipeline tremor://localhost/pipeline/system::metrics/system/in
...
...
{"onramp":"metronome","ingest_ns":1621942260121807140,"id":486}
{"onramp":"metronome","ingest_ns":1621942270121987716,"id":487}
{"onramp":"metronome","ingest_ns":1621942280122466690,"id":488}
{"onramp":"metronome","ingest_ns":1621942290101218178,"id":489}
{"onramp":"metronome","ingest_ns":1621942300101332444,"id":490}If you can see metronome events with the default example then your deployment looks good!
In case something goes wrong you can always check events to see what's going on.
$ kubectl get events --sort-by=.metadata.creationTimestamp -n tremor-test
LAST SEEN TYPE REASON OBJECT MESSAGE
4m16s Normal SuccessfulCreate replicaset/tremor-test-84f659495d Created pod: tremor-test-84f659495d-5v759
4m16s Normal Scheduled pod/tremor-test-84f659495d-5v759 Successfully assigned tremor-test/tremor-test-84f659495d-5v759 to gke-44c3ebc6-2nqu
4m16s Normal ScalingReplicaSet deployment/tremor-test Scaled up replica set tremor-test-84f659495d to 1
4m14s Normal Pulled pod/tremor-test-84f659495d-5v759 Container image "tremorproject/tremor:0.11.4" already present on machine
4m14s Normal Created pod/tremor-test-84f659495d-5v759 Created container tremor
4m14s Normal Started pod/tremor-test-84f659495d-5v759 Started container tremorBy default HPA is disabled. Once you enable it you can check configuration:
$ kubectl get hpa -n tremor-test
NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
tremor-test Deployment/tremor-test 5%/50%, 4%/50% 1 2 1 3h6m$ kubectl describe hpa.v2beta2.autoscaling tremor-test -n tremor-test
Name: tremor-test
Namespace: tremor-test
Labels: app=tremor-test
chart=tremor-0.1.0-rc.5.b4b8b7851
component=tremor
heritage=Tiller
release=tremor-test
Annotations: <none>
CreationTimestamp: Thu, 03 Jun 2021 11:46:08 +0200
Reference: Deployment/tremor-test
Metrics: ( current / target )
resource memory on pods (as a percentage of request): 5% (14409728) / 50%
resource cpu on pods (as a percentage of request): 4% (4m) / 50%
Min replicas: 1
Max replicas: 2
Deployment pods: 1 current / 1 desired
Conditions:
Type Status Reason Message
---- ------ ------ -------
AbleToScale True ReadyForNewScale recommended size matches current size
ScalingActive True ValidMetricFound the HPA was able to successfully calculate a replica count from memory resource utilization (percentage of request)
ScalingLimited False DesiredWithinRange the desired count is within the acceptable range
Events: <none>
We use github pages to host a generated helm chart repository for helm chart releases here.