tfmodule-azure-redis-enterprise

This github uses anisble and terraform to create a full redis cluster of a specified number of nodes. Optionally, additional one node tiers are available: cassandra, ubuntu application node, and kafka node.

Full Deployment

Prerequisites

• azure-cli
• tfenv

Initialize Azure Credentials

az login

Links

installing, configuring, running confluent platform Azure terraform provider

Module usage (Basic)

Public Azure Instance and Network setup. Note the count variables:

• node-count: number of redis nodes (3 is most common)
• cassandra-count: number of cassandra nodes (really just 1 or 0)
• tester-count: number of application nodes (really just 1 or 0)
• kafka-count: number of kafka nodes (really just 1 or 0)

To allow access from local PC instigating the terraform/ansible, set the my-ip variable. This allows connectivity to redis and cassandra ports from the instigating laptop IP address. Easiest way to get this Azure public IP address is using the Azure portal. Go to an existing security rule in the portal and use the My IP address option to see local desktop IP address to put in to the my-ip variable.

Module usage (More advanced)

module "azure-redis-enterprise" {
  source       = "git@github.com:Redislabs-Solution-Architects/tfmodule-azure-redis-enterprise.git"
  cluster-name = "terraform-test-1.azure.example.com"
  location     = "uswest2"
  common-tags = {
    Owner       = "Ken_Watanabe@example.com",
    Config      = "terraform",
    Environment = "tf-test"
  }
}

From the test folder, update the main.tf file with the correct parameters for your environment.

Run

terraform init

terraform plan

terraform apply

Check out what is there

node addresses for cassandra and kafka, as well as the redis endpoints and passord are dumped as appropriately named text files in the directory provisioners/temp

kafka node

Check out the kafka control panel by substituting the localhost with the ip found in kafka_external_ip.txt at http://localhost:9021

cassandra

export CQLSH_HOST=<public ip from cassandra_external_ip.txt>
cqlsh
> describe keyspaces;
pageviews  system_auth         system_schema  system_views
system     system_distributed  system_traces  system_virtual_schema

>use pageviews;
>select * from views;
 viewtime | pageid  | userid
----------+---------+--------
   123791 | Page_66 | User_4
   619431 | Page_99 | User_5
   972821 | Page_20 | User_8
   437401 | Page_51 | User_4
    26941 | Page_39 | User_4
  1021611 | Page_62 | User_5
   564921 | Page_13 | User_2
   504541 | Page_23 | User_4
   247411 | Page_20 | User_6
   463431 | Page_70 | User_7
   733721 | Page_55 | User_6
   376961 | Page_75 | User_2
   726761 | Page_68 | User_8
   242211 | Page_22 | User_9
   271671 | Page_55 | User_4
   388351 | Page_85 | User_4

redis

redis-cli -h <redis_external_endpoint.txt> -p <redis_port.txt> -a redis123
6077) "pageviews:27061"
6078) "pageviews:65441"
6079) "pageviews:60781"
6080) "pageviews:3191"
6081) "pageviews:36801"
6082) "pageviews:33691"
6083) "pageviews:7971"
6084) "pageviews:24301"
6085) "pageviews:35591"

Tech notes

• test/main.tf has the important parameters.
• Other paramaters are available in variables.tf.
• A template file at templates/inventory.tpl maps connects terraform to ansible along with provisioning.tf
• Ansible is all in the provisioners directory using the playbook.yml file
• Each of the node groups (redis, tester, cassandra, kafka) have a separate role under the provisioners/role directory (cassandra-node, kafka-node, redis-enterprise, tester-node)
• Under each of these roles, a vars/main.yml file has variable flags to enable/disable processing
  • To eliminate all process on one of these node groups, best to set the node count to zero in test/main.tf
• Under each of these roles a tasks/main.yml calls the required tasks to do the actual processing

To tear it all down: NOTE: on teardown, may see failures on delete of some azure components. Re-running the destroy command will eventually be successful

terraform destroy