aws-terraform-elasticache

This module creates Elasticache-Memcached, Elasticache-Redis, or Elasticache-Redis Multi-Shard instances.

Basic Usage

module "elasticache_memcached" {
  source = "git@github.com:rackspace-infrastructure-automation/aws-terraform-elasticache.git?ref=v0.12.0"

  create_internal_zone_record = true
  curr_connections_threshold  = 500
  elasticache_engine_type     = "memcached14"
  evictions_threshold         = 10
  instance_class              = "cache.m4.large"
  internal_record_name        = "memcachedconfig"
  internal_zone_id            = module.internal_zone.internal_hosted_zone_id
  internal_zone_name          = module.internal_zone.internal_hosted_name
  name                        = "memc-${random_string.name_suffix.result}"
  security_groups             = [module.security_groups.elastic_cache_memcache_security_group_id]
  subnets                     = module.vpc.private_subnets

  tags = {
    MyTag1 = "MyValue1"
    MyTag2 = "MyValue2"
    MyTag3 = "MyValue3"
  }
}
*

Full working references are available at examples

Other TF Modules Used

Using aws-terraform-cloudwatch_alarm to create the following CloudWatch Alarms:

• evictions_alarm
• cpu_utilization_alarm
• curr_connections_alarm
• swap_usage_alarm

Terraform 0.12 upgrade

Several changes were required while adding terraform 0.12 compatibility. The following changes should be
made when upgrading from a previous release to version 0.12.0 or higher.

Module variables

The following module variables were updated to better meet current Rackspace style guides:

• additional_tags -> tags
• cluster_name -> name
• cluster_name_version -> name_version
• create_route53_record -> create_internal_zone_record
• security_group_list -> security_groups

Requirements

Name	Version
terraform	>= 0.12
aws	>= 2.7.0

Providers

Name	Version
aws	>= 2.7.0
null	n/a

Modules

Name	Source	Version
cpu_utilization_alarm	git@github.com:rackspace-infrastructure-automation/aws-terraform-cloudwatch_alarm//?ref=v0.12.6
curr_connections_alarm	git@github.com:rackspace-infrastructure-automation/aws-terraform-cloudwatch_alarm//?ref=v0.12.6
evictions_alarm	git@github.com:rackspace-infrastructure-automation/aws-terraform-cloudwatch_alarm//?ref=v0.12.6
swap_usage_alarm	git@github.com:rackspace-infrastructure-automation/aws-terraform-cloudwatch_alarm//?ref=v0.12.6

Resources

Name
aws_elasticache_cluster
aws_elasticache_parameter_group
aws_elasticache_replication_group
aws_elasticache_subnet_group
aws_route53_record
null_data_source

Inputs

Name	Description	Type	Default	Required
at_rest_encrypted_disk	(redis, redis multi shard) Indicates whether to enable encryption at rest. ONLY AVAILABLE FOR REDIS 3.2.6, 4.0.10 AND 5.0.0. true or false.	bool	false	no
authentication_token	(redis, redis multi shard) The password used to access a password protected server. Can be specified only if in_transit_encryption = true and will be ignored otherwise	string	""	no
cache_cluster_port	(memcached, redis, redis multi shard) The port number on which each of the cache nodes will accept connections. Default for redis is 6379. Default for memcached is 11211	string	""	no
cluster_name	(memcached, redis, redis multi shard) Name of the Cluster. Will also be used to name other provisioned resources. If this value is empty, name and name_version will be used for backwards compatibility. Should not exceed 50 characters due to string length constraints. NOTICE: Cluster Name allows 50 characters MAX.	string	""	no
cpu_high_evaluations	(memcached, redis) The number of minutes CPU usage must remain above the specified threshold to generate an alarm.	number	5	no
cpu_high_threshold	(memcached, redis) The max CPU Usage % before generating an alarm.	number	90	no
create_internal_zone_record	(memcached, redis, redis multi shard) Specifies whether or not to create a route53 CNAME record for the configuration/primary endpoint. internal_zone_id, internal_zone_name, and internal_record_name must be provided if set to true. true or false.	bool	false	no
curr_connections_evaluations	(memcached, redis) The number of minutes current connections must remain above the specified threshold to generate an alarm.	number	5	no
curr_connections_threshold	(memcached, redis) The max number of current connections before generating an alarm. NOTE: If this variable is not set, the connections alarm will not be provisioned.	string	""	no
elasticache_engine_type	(memcached, redis, redis multi shard) The name of the cache engine to be used for this cluster. Valid values are: memcached14, memcached1510, memcached1516, redis28, redis2823, redis2822, redis2821, redis2819, redis286, redis32, redis326, redis3210, redis40, redis50, redis503, redis504, redis505, redis506	string	n/a	yes
environment	(memcached, redis, redis multi shard) Application environment for which this network is being created. Preferred value are Development, Integration, PreProduction, Production, QA, Staging, or Test	string	"Development"	no
evictions_evaluations	(memcached, redis) The number of minutes Evictions must remain above the specified threshold to generate an alarm.	number	5	no
evictions_threshold	(memcached, redis) The max evictions before generating an alarm. NOTE: If this variable is not set, the evictions alarm will not be provisioned.	string	""	no
failover_enabled	(redis) Enable Multi-AZ Failover. Failover is unsupported on the cache.t1.micro instance class. This is hardcoded as true for Redis multi-shard.	bool	true	no
in_transit_encryption	(redis, redis multi shard) Indicates whether to enable encryption in transit. Because there is some processing needed to encrypt and decrypt the data at the endpoints, enabling in-transit encryption can have some performance impact.ONLY AVAILABLE FOR REDIS 3.2.6, 4.0.10 and later. true or false	bool	false	no
instance_class	(memcached, redis, redis multi shard) The compute and memory capacity of the nodes within the ElastiCache cluster. Please see https://docs.aws.amazon.com/AmazonElastiCache/latest/mem-ug/CacheNodes.SupportedTypes.html for valid instance types.	string	n/a	yes
internal_record_name	(memcached, redis, redis multi shard) Record Name for the new Resource Record in the Internal Hosted Zone	string	""	no
internal_zone_id	(memcached, redis, redis multi shard) The Route53 Internal Hosted Zone ID	string	""	no
internal_zone_name	(memcached, redis, redis multi shard) LD for Internal Hosted Zone	string	""	no
name	DEPRECATED, MAY REMOVE IN A FUTURE VERSION, USE cluster_name AND replication_group_name!! (memcached, redis, redis multi shard) Name of Cluster. Will also be used to name other provisioned resources. If non empty cluster_name_version is being used, total length of cluster_name plus cluster_name_version should not exceed 19 due to string length constraints	string	""	no
name_version	DEPRECATED, MAY REMOVE IN A FUTURE VERSION, USE cluster_name AND replication_group_name!! (memcached, redis, redis multi shard) NOTE: This needs to increment on update with new snapshot. If non empty cluster_name_version is being used, total length of cluster_name plus cluster_name_version should not exceed 19 due to string length constraints	string	"v00"	no
notification_topic	(memcached, redis, redis multi shard) SNS Topic ARN to notify if there are any alarms	string	""	no
number_of_nodes	(memcached, redis) The number of cache nodes within the ElastiCache cluster. This number must be grearter or equal 2 to enable automatic failover.	number	1	no
number_of_read_replicas_per_shard	(redis multi shard) Number of read replicas per shard	number	2	no
number_of_shards	(redis multi shard) Number of shards	number	2	no
preferred_maintenance_window	(memcached, redis, redis multi shard) The weekly time range (in UTC) during which system maintenance can occur. Example: Sun:05:00-Sun:07:00	string	"Sun:05:00-Sun:07:00"	no
redis_multi_shard	(edis, redis multi shard) Is this a redis multi-shard instance? true or false	bool	false	no
replication_group_description	(redis, redis multi shard) Description of Replication Group	string	"Elasticache"	no
replication_group_name	(memcached, redis, redis multi shard) Name of the Cluster replication group ID, must be unique. If this value is empty, name and name_version will be used for backwards compatibility. Should not exceed 40 characters due to string length constraints. NOTICE: Group ID allows 40 characters MAX.	string	""	no
security_groups	(memcached, redis, redis multi shard) A list of EC2 security groups to assign to this resource.	list(string)	n/a	yes
snapshot_arn	(redis, redis multi shard) The S3 ARN of a snapshot to use for cluster creation. Proper access to the S3 file must be granted prior to building instance. See https://docs.aws.amazon.com/AmazonElastiCache/latest/UserGuide/backups-seeding-redis.html#backups-seeding-redis-grant-access for details. This parameter is ignored if a SnapshotName is provided.	string	""	no
snapshot_name	(redis, redis multi shard) The name of a snapshot to use for cluster creation. This property overrides any value assigned to SnapshotArn. Snapshots are unsupported on cache.t1.micro instance class.	string	""	no
snapshot_retention_limit	(redis, redis multi shard) The number of days for which automated backups are retained. Setting this parameter to a positive number enables backups. Setting this parameter to 0 disables automated backups. Snapshots are unsupported on cache.t1.micro instance class.	number	7	no
snapshot_window	(redis, redis multi shard) The daily time range (in UTC) during which ElastiCache will begin taking a daily snapshot of your node group. Snapshots are unsupported on cache.t1.micro instance class.	string	"03:00-05:00"	no
subnets	(memcached, redis, redis multi shard) List of subnets for use with this cache cluster	list(string)	n/a	yes
swap_usage_evaluations	(memcached) The number of minutes SwapUsage must remain above the specified threshold to generate an alarm	number	5	no
swap_usage_threshold	(memcached) The max SwapUsage before generating an alarm	number	52428800	no
tags	(memcached, redis, redis multi shard) Additional tags to be added to the Elasticache resources. Please see examples directory in this repo for examples.	map(string)	{}	no

Outputs

Name	Description
elasticache_endpoint	Elasticache endpoint address
elasticache_internal_r53_record	Internal Route 53 record FQDN for the Elasticache endpoint(s)