Apache SOLR Cloud Helm Chart

This helm chart provides an implementation of Apache Solr StatefulSet.

• Apache SOLR Cloud Helm Chart
  • Prerequisites
  • Chart Components
  • Installing the Chart
    • Installed Components
  • Configuration
  • Default Values
  • Deep Dive
    • Images
      • Zookeeper Image
      • Solr Image
    • Scaling
      • Solr
      • Zookeeper
    • Access API and Client Searching
    • Create a Collection
    • Add Data to Collection
    • Search Data in Collection

Prerequisites

• Kubernetes 1.8+
• PersistentVolume support
• A familiarity with Apache Solr 7
• Helm Chart Zookeeper 1.2.0 see requirements.yaml

Chart Components

This chart will do the following:

• Create three (3) Solr Cloud nodes in a Statefulset
• Create three (3) Zookeeper zNodes in an ensemble.
• Create a Headless Service to control the Solr Cloud replicas.
• Create a Service configured to perform Admin functions and allow client connections

Installing the Chart

You can install the chart with the release name myrelease as below.

helm repo add incubator http://storage.googleapis.com/kubernetes-charts-incubator
helm upgrade myrelease --install incubator/solr

Note - If you do not specify a name, helm will select a name for you.

Installed Components

You can use kubectl get to view all the installed components.

kubectl get all -l release=myrelease
NAME                               DESIRED   CURRENT   AGE
statefulsets/myrelease-solr        3         3         4m
statefulsets/myrelease-zookeeper   3         3         4m

NAME                       READY     STATUS    RESTARTS   AGE
po/myrelease-solr-0        1/1       Running   0          4m
po/myrelease-solr-1        1/1       Running   0          3m
po/myrelease-solr-2        1/1       Running   0          2m
po/myrelease-zookeeper-0   1/1       Running   0          4m
po/myrelease-zookeeper-1   1/1       Running   0          4m
po/myrelease-zookeeper-2   1/1       Running   0          3m

NAME                               TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                      AGE
svc/myrelease-solr                 ClusterIP   10.110.148.187   <none>        80/TCP                       4m
svc/myrelease-solr-headless        ClusterIP   None             <none>        8983/TCP                     4m
svc/myrelease-zookeeper            ClusterIP   10.108.231.151   <none>        2181/TCP                     4m
svc/myrelease-zookeeper-headless   ClusterIP   None             <none>        2181/TCP,3888/TCP,2888/TCP   4m

Configuration

You can specify each parameters using the --set key=value argument either in helm install or helm upgrade

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

helm install --name myrelease -f values.yaml incubator/solr

helm upgrade --install myrelease -f values.yaml incubator/solr

Default Values

The configuration parameters in this section control the resources requested and utilized by the Apache SOLR Cloud Instance.

Parameter	Description	Default
replicaCount	Number of SOLR Cloud Replicas to install or scale too	3
revisionHistoryLimit	Number of Revision History to keep. See Docs	2
updateStrategy.type	updateStrategy for the StatefulSet	RollingUpdate
podManagementPolicy	Pod's Management Policy See Docs
terminationGracePeriodSeconds	Terminate the pod after x seconds if hook fails	1800
ports.client.containerPort	Container port for SOLR Cloud	8983
heap	Apache SOLR Heap Settings	2g
timeZone	Apache SOLR TimeZone Settings	UTC
LogLevel	Apache SOLR Log Level Settings	INFO
image.repository	Docker Hub Library Apache SOLR Image	solr
image.tag	Docker Hub Library Apache Solr Image Tag	7.5
image.pullPolicy	Image Pull Policy	IfNotPresent
service.type	Client Service Type	ClusterIP
service.port	Client Service Port	80
ingress.enabled	Ingress Enabled	true
ingress.annotations	Ingress Annotations	nil
ingress.path	Ingress Path	/
ingress.hosts	Hosts used for Ingress	solr-admin.local
persistence.enabled	Enabled Persistence	true
persistence.storageClass	Storage Class to be used	Commented Out
persistence.accessMode	Data Access Mode to be used for the Data Directory	ReadWriteOnce
persistence.size	PVC Size for Data Directory	1Gi

Notes - Read through the Zookeeper Default Values to customize the Zookeeper instance

Deep Dive

Images

Zookeeper Image

This chart requires a Zookeeper ensemble to start. The image used for this is current in the helm chart incubator.

Solr Image

The image used for this chart is based on Apache Solr 7.5.0 in the Docker Hub library for SOLR.

Scaling

Solr

You can scale the number of replica for Solr by changing the replicaCount either in the values.yaml or the set replicaCount via the helm install or helm upgrade command. Scaling the number of replicas DOES NOT add replicas to an existing collection. This will have to be done manually via a CI/CD build process.

Zookeeper

Read the section on Scaling Zookeeper in the Helm Chart Documentation for Zookeeper. It is recommended that you set a fixed number of Zookeeper znodes using the zookeeper.replicaCount in the values.yaml. We have found inconsistencies with Solr when scaling Zookeeper after the chart has been deployed.

Access API and Client Searching

If you enabled ingress, users and administrators can search and perform administrative tasks on the SOLR Cloud replicas or collection(s). You can access the instance via the host name you supplied in the values.yaml. The default is solr-admin.local. You will see information in the NOTES.txt on how to access the URL or after you install the chart.

Note - If you are working locally you will need to edit your /etc/hosts file to include an entry for solr-admin.local.

• Example - Notes.txt output after install

NOTES:
1. View the SOLR Cloud GUI via the url:
  http://solr-admin.local/

Create a Collection

You can create a SOLR Collection using the following command. This will create one (1) collection called testcoll and one (1) shard with a replicationFactor of three (3).

kubectl run --attach adminbox --image=centos --restart=Never -- sh -c 'curl -X POST "http://myrelease-solr-headless:8983/solr/admin/collections?action=CREATE&name=testcoll&numShards=1&replicationFactor=3"';
If you don't see a command prompt, try pressing enter.
100   859  100   859    0     0    145      0  0:00:05  0:00:05 --:--:--   220
{
  "responseHeader":{
    "status":0,
    "QTime":5893},
  "success":{
    "myrelease-solr-1.myrelease-solr-headless:8983_solr":{
      "responseHeader":{
        "status":0,
        "QTime":3539},
      "core":"testcoll_shard1_replica_n4"},
    "myrelease-solr-2.myrelease-solr-headless:8983_solr":{
      "responseHeader":{
        "status":0,
        "QTime":3776},
      "core":"testcoll_shard1_replica_n3"},
    "myrelease-solr-0.myrelease-solr-headless:8983_solr":{
      "responseHeader":{
        "status":0,
        "QTime":3939},
      "core":"testcoll_shard1_replica_n1"}},
  "warning":"Using _default configset. Data driven schema functionality is enabled by default, which is NOT RECOMMENDED for production use. To turn it off: curl http://{host:port}/solr/testcoll/config -d '{\"set-user-property\": {\"update.autoCreateFields\":\"false\"}}'"}

Add Data to Collection

You can add data to the collection via the ingress.

curl http://solr-admin.local/solr/testcoll/update -H "Content-Type: text/xml" --data-binary '
<add>
  <doc>
    <field name="authors">Patrick Eagar</field>
    <field name="subject">Sports</field>
    <field name="dd">796.35</field>
    <field name="isbn">0002166313</field>
    <field name="yearpub">1982</field>
    <field name="publisher">Collins</field>
  </doc>
</add>'

curl "http://solr-admin.local/solr/testcoll/update" --data '<commit/>'

Search Data in Collection

You can search data in the collecation via ingress.

curl "http://solr-admin.local/solr/testcoll/select?q=*:*"