Birdhouse Helm Chart

This Helm Chart installs some of the components of the Birdhouse project into a Kubernetes cluster. In this deployment Phoenix acts as the WPS web-based client to Emu, Hummingbird, FlyingPigeon or other WPS you may want to deploy and register into Phoenix.

TL;DR

git clone https://github.com/Pablo-CC/birdhouse-helm.git
kubectl create namespace birdhouse
helm install --name birdhouse --namespace birdhouse .

Configuration

Parameter	Description	Default
phoenix.name	The name for the phoenix deployment	phoenix
phoenix.image	The image for the phoenix container	birdhouse/pyramid-phoenix
phoenix.imagePullPolicy	The image pull policy	IfNotPresent
phoenix.replicas	The number of Phoenix's replicas for the deployment to build	1
phoenix.hostname	The FQDN of Phoenix [*]	phoenix.xxx.xxx.xxx.xxx.nip.io
phoenix.ports.http	The port to listen on for HTTP requests	8081
phoenix.ports.https	The port to listen on for HTTPS requests	8443
phoenix.mongodb.name	The name for the deployment of the MongoDB instance	mongodb
phoenix.mongodb.port	The port that MongoDB listens on	27017
phoenix.mongodb.image	The image for the MongoDB container	mongo:2.6.12
phoenix.mongodb.imagePullPolicy	The image pull policy	IfNotPresent
phoenix.resources.requests.memory	The minimum memory requests for Phoenix (required for HPA on Memory usage)	1Gi
phoenix.resources.requests.cpu	The minimum CPU request for Phoenix (required for HPA on CPU usage)	400m
phoenix.hpa.enabled	Boolean for enabling Horizontal Pod Autoscaling for Phoenix	true
phoenix.hpa.minReplicas	The minimum number of replicas for the Autoscaler	1
phoenix.hpa.maxReplicas	The maximum number of replicas for the Autoscaler	8
phoenix.hpa.cpuUtilization	Autoscaling trigger as a percentage of phoenix.resources.requests.cpu	80
phoenix.hpa.memUtilization	Autoscaling trigger as a percentage of phoenix.resources.requests.memory	100
emu.enabled	Boolean for enabling EMU as a backend Web Processing Service	true
emu.name	The name for the Emu Deployment	emu
emu.image	The image for Emu's container	birhouse/emu
emu.imagePullPolicy	The image pull policy	IfNotPresent
emu.replicas	The number of Emu's replicas for the deployment to build	1
emu.hostname	The FQDN of Emu [*]	emu.xxx.xxx.xxx.xxx.nip.io
emu.port	The port on which Emu listens	5000
emu.logLevel	The log level for Emu [DEBUG|INFO|WARNING|ERROR|CRITICAL|FATAL]	WARNING
hummingbird.enabled	Boolean for enabling Hummingbird as a backend Web Processing Service	true
hummingbird.name	The name for the Hummingbird deployment	hummingbird
hummingbird.image	The image for Hummingbird's container	birdhouse/hummingbird
hummingbird.imagePullPolicy	The image pull policy	IfNotPresent
hummingbird.replicas	The number of Hummingbird's replicas for the deployment to build	1
hummingbird.hostname	The FQDN of FlyingPigeon	hummingbird.xxx.xxx.xxx.xxx.nip.io
hummingbird.port	The port on which Hummingbird listens	5000
hummingbird.logLevel	The log level for Emu [DEBUG|INFO|WARNING|ERROR|CRITICAL|FATAL]	WARNING
flyingpigeon.enabled	Boolean for enabling FlyingPigeon as a backend Web Processing Service	true
flyingpiugeon.name	The name for the FlyingPigeon deployment	flyingpigeon
flyingpigeon.image	The image for FlyingPigeon's container	birdhouse/flyingpigeon
flyingpigeon.imagePullPolicy	The image pull policy	IfNotPresent
flyingpigeon.replicas	The number of FlyingPigeon's replicas for the deployment to build	1
flyingpigeon.hostname	The FQDN of FlyingPigeon [*]	flyingpigeon.xxx.xxx.xxx.xxx.nip.io
flyingpigeon.port	The port on which FlyingPigeon listens	5000
flyingpigeon.logLevel	The log level for FlyingPigeon [DEBUG|INFO|WARNING|ERROR|CRITICAL|FATAL]	WARNING

[*] Tip: You can use nip.io

Specify each parameter using the --set key=value[,key=value] argument to helm install. For example,

$ helm install --name my-release --namespace birdhouse --set emu.enabled=false .

Alternatively, a YAML file that specifies the values for the parameters can be provided while installing the chart. For example,

$ helm install --name my-release -f my_values.yaml .

Autoscaling

Kubernetes' Horizontal Pod Autoscaler is used for the autoscaling of the frontend (i.e. Phoenix). The following figures represent the results of this during a load test:

CPU Usage

Memory Usage

Replicas + Requests

Replicas

Requests per second