Terraform module that configures the recommended Amazon ElasticSearch Alarms using CloudWatch and sends alerts to an SNS topic. By default, this module creates an SNS topic, but it can be configured to point to an existing SNS topic (see example)
v1.x supports terraform v0.12+ syntax!
v2.0+ supports terraform v1.0+ syntax!
This project is inspired by CloudPosse
It's 100% Open Source and licensed under the APACHE2.
| Area | Metric | Operator | Threshold | Rationale |
|---|---|---|---|---|
| Sharding | ClusterStatus.red | >= |
1 | At least one primary shard and its replicas are not allocated to a node |
| Sharding | ClusterStatus.yellow | >= |
1 | At least one replica shard is not allocated to a node |
| Storage | FreeStorageSpace | <= |
20480 MB | A node in your cluster is down to low storage space. Note, this alarm uses the aggregate Minimum which means this alarm triggers per-node in your cluster. This logic is based-on the AWS Recommended Alarms. It does not however alarm based on an aggregate of free space remaining. |
| Storage | FreeStorageSpaceTotal | <= |
20480 MB | The overall disk space free is low. This alarm uses Sum across all your nodes, this can be useful on multi-node clusters. Disabled by default, to enable this you must set monitor_free_storage_space_total_too_low to true, and free_storage_space_total_threshold. Recommended to set the threshold to the number of nodes in your cluster multiplied by the free_storage_space_threshold |
| Storage | ClusterIndexWritesBlocked | >= |
1 | Your cluster is blocking write requests. |
| Node Count | Nodes | < |
x |
This alarm indicates that at least one node in your cluster has been unreachable for one day |
| Snapshot | AutomatedSnapshotFailure | >= |
1 | An automated snapshot failed. This failure is often the result of a red cluster health status. |
| CPU | CPUUtilization | >= |
80 % | 100% CPU utilization isn't uncommon, but sustained high usage is problematic. Consider using larger instance types or more instances. |
| Memory | JVMMemoryPressure | >= |
80 % | The cluster could encounter out of memory errors if usage increases. Consider scaling vertically. |
| CPU | MasterCPUUtilization | >= |
80 % | Consider using larger instance types for your dedicated master nodes. |
| Memory | MasterJVMMemoryPressure | >= |
80 % | Consider using larger instance types for your dedicated master nodes. |
| KMS | KMSKeyError | >= |
1 | The KMS encryption key that is used to encrypt data at rest in your domain is disabled. Re-enable it to restore normal operations |
| Memory | KMSKeyInaccessible | >= |
80 % | The KMS encryption key that is used to encrypt data at rest in your domain has been deleted or has revoked its grants to Amazon ES |
For more information please see: recommended Amazon ElasticSearch Alarms.
See the examples/ directory for working examples.
resource "aws_elasticsearch_domain" "es" {
domain_name = "example"
elasticsearch_version = "7.10"
cluster_config {
instance_type = "r4.large.elasticsearch"
}
snapshot_options {
automated_snapshot_start_hour = 23
}
tags = {
Domain = "TestDomain"
}
}
module "es_alarms" {
source = "github::https://github.com/dubiety/terraform-aws-elasticsearch-cloudwatch-sns-alarms.git?ref=master"
domain_name = "example"
tags = {
Domain = "TestDomain"
}
}You can alternatively have this module not create an SNS in case you have existing ones created elsewhere.
module "es_alarms" {
source = "github::https://github.com/dubiety/terraform-aws-elasticsearch-cloudwatch-sns-alarms.git?ref=master"
domain_name = "example"
sns_topic = "arn:aws:sns:us-east-1:123456123456:sns-to-slack" # < Put your full SNS ARN here, if necessary read from var or a resource
create_sns_topic = false
tags = {
Domain = "TestDomain"
}
}| Name | Description | Type | Default | Required |
|---|---|---|---|---|
domain_name |
The Elasticserach domain name you want to monitor. | string | - | yes |
cluster_type |
The type of cluster, single or multi-node | string | "single" |
no |
alarm_name_postfix |
Alarm name postfix | string | "" |
no |
alarm_name_prefix |
Alarm name prefix | string | "" |
no |
create_sns_topic |
Will create an SNS topic, if you set this to false you MUST set sns_topic to a FULL ARN |
bool | true |
no |
sns_topic |
SNS topic you want to specify. If leave empty, it will use a prefix and a timestamp appended. If create_sns_topic is set to false, this MUST be a FULL ARN |
string | "" |
no |
sns_topic_postfix |
SNS topic postfix | string | "" |
no |
sns_topic_prefix |
SNS topic prefix | string | "" |
no |
tags |
Tags to associate with all created resources | map | {} |
no |
cpu_utilization_threshold |
The maximum percentage of CPU utilization | string | 80 |
no |
free_storage_space_threshold |
The minimum amount of available storage space in MiB. | string | 20480 |
no |
jvm_memory_pressure_threshold |
The maximum percentage of the Java heap used for all data nodes in the cluster | string | 80 |
no |
master_cpu_utilization_threshold |
The maximum percentage of CPU utilization of master nodes | string | "" |
no |
master_jvm_memory_pressure_threshold |
The maximum percentage of the Java heap used for master nodes in the cluster | string | "" |
no |
min_available_nodes |
The minimum available (reachable) nodes to have, set to non-zero to enable alarm | string | 0 |
no |
monitor_automated_snapshot_failure |
Enable monitoring of automated snapshot failure | bool | true |
no |
monitor_cluster_status_is_red |
Enable monitoring of cluster status is in red | bool | true |
no |
monitor_cluster_status_is_yellow |
Enable monitoring of cluster status is in yellow | bool | true |
no |
monitor_cluster_index_writes_blocked |
Enable monitoring of cluster index writes being blocked | bool | true |
no |
monitor_cpu_utilization_too_high |
Enable monitoring of CPU utilization is too high | bool | true |
no |
monitor_free_storage_space_too_low |
Enable monitoring of minimum per-node free storage is too low | bool | true |
no |
monitor_free_storage_space_total_too_low |
Enable monitoring of cluster total free storage is too low | bool | false |
no |
monitor_jvm_memory_pressure_too_high |
Enable monitoring of JVM memory pressure is too high | bool | true |
no |
monitor_kms |
Enable monitoring of KMS-related metrics, enable if using KMS | bool | false |
no |
monitor_master_cpu_utilization_too_high |
Enable monitoring of CPU utilization of master nodes are too high. Only enable this when dedicated master is enabled | bool | false |
no |
monitor_master_jvm_memory_pressure_too_high |
Enable monitoring of JVM memory pressure of master nodes are too high. Only enable this wwhen dedicated master is enabled | bool | false |
no |
monitor_min_available_nodes |
Enable monitoring of minimum available nodes | bool | true |
no |
alarm_automated_snapshot_failure_periods |
The number of periods to alert that automatic snapshots failed, raise this if desired to make less noisy | number | 1 |
no |
alarm_cluster_status_is_red_periods |
The number of periods to alert that cluster status is red, raise this to be less noisy | number | 1 |
no |
alarm_cluster_status_is_yellow_periods |
The number of periods before triggering the cluster status is yellow, raise this to be less noisy | number | 1 |
no |
alarm_cluster_index_writes_blocked_periods |
The number of periods to alert that cluster index writes are blocked, raise this if desired to make less noisy | number | 1 |
no |
alarm_cpu_utilization_too_high_periods |
The number of periods to alert that CPU usage is too high, raise this if desired to make less noisy | number | 3 |
no |
alarm_free_storage_space_too_low_periods |
The number of periods before triggering the disk space is low, raise this to be less noisy | number | 1 |
no |
alarm_free_storage_space_total_too_low_periods |
The number of periods before triggering the total disk space is low, raise this to be less noisy | number | 1 |
no |
alarm_jvm_memory_pressure_too_high_periods |
The number of periods which it must be in the alarmed state to alert, raise this if desired to make less noisy | number | 1 |
no |
alarm_kms_periods |
The number of periods to alert that kms has failed, raise this if desired to make less noisy | number | 1 |
no |
alarm_master_cpu_utilization_too_high_periods |
The number of periods to alert that masters CPU usage is too high, raise this if desired to make less noisy | number | 3 |
no |
alarm_master_jvm_memory_pressure_too_high_periods |
The number of periods which it must be in the alarmed state to alert, raise this if desired to make less noisy | number | 1 |
no |
alarm_min_available_nodes_periods |
The number of periods to alert that minimum number of available nodes dropped below a threshold, raise this if desired to make less noisy | number | 1 |
no |
alarm_min_available_nodes_period |
The period of the minimum available nodes should the statistics be applied in seconds | string | 86400 |
no |
alarm_automated_snapshot_failure_period |
The period of the automated snapshot failure should the statistics be applied in seconds | string | 60 |
no |
alarm_cluster_index_writes_blocked_period |
The period of the cluster index writes being blocked should the statistics be applied in seconds | string | 300 |
no |
alarm_cluster_status_is_red_period |
The period of the cluster status is in red should the statistics be applied in seconds | string | 60 |
no |
alarm_cluster_status_is_yellow_period |
The period of the cluster status is in yellow should the statistics be applied in seconds | string | 60 |
no |
alarm_cpu_utilization_too_high_period |
The period of the CPU utilization is too high should the statistics be applied in seconds | string | 900 |
no |
alarm_free_storage_space_too_low_period |
The period of the per-node minimum free storage is too low should the statistics be applied in seconds | string | 60 |
no |
alarm_free_storage_space_total_too_low_period |
The period of the cluster total free storage is too low should the statistics be applied in seconds | string | 60 |
no |
alarm_jvm_memory_pressure_too_high_period |
The period of the JVM memory pressure is too high should the statistics be applied in seconds | string | 900 |
no |
alarm_kms_period |
The period of the KMS-related metrics should the statistics be applied in seconds | string | 60 |
no |
alarm_master_cpu_utilization_too_high_period |
The period of the CPU utilization of master nodes are too high should the statistics be applied in seconds | string | 900 |
no |
alarm_master_jvm_memory_pressure_too_high_period |
The period of the JVM memory pressure of master nodes are too high should the statistics be applied in seconds | string | 900 |
no |
| Name | Description |
|---|---|
sns_topic_arn |
The ARN of the SNS topic |
sns_topic_name |
The SNS topic name |
Like this project? Please give it a ★ on our GitHub!
Got a question?
File a GitHub issue.
Please use the issue tracker to report any bugs or file feature requests.
See LICENSE for full details.
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.