Database Monitoring은 중요한 메트릭, 작업 지연, 작업 샘플, Explain 실행 계획 및 복제 상태 변경 정보에 액세스할 수 있도록 지원하며, 이를 통해 MongoDB 데이터베이스에 대한 포괄적인 인사이트를 제공합니다. MongoDB용 Database Monitoring을 활용하려면 Datadog Agent가 설치되어 있고 MongoDB Atlas 인스턴스에 연결되도록 구성되어 있는지 확인하세요. 이 가이드에서는 Database Monitoring을 설정하는 단계를 간략하게 설명합니다.
시작 전 참고 사항
지원되는 MongoDB 주요 버전
4.4, 5.0, 6.0, 7.0, 8.0
지원되는 MongoDB Atlas 클러스터 티어
M10 이상
참고: MongoDB Atlas Serverless 인스턴스 또는 공유 클러스터(M0 Sandbox, M2, M5)는 지원되지 않습니다.
Supported Agent versions
7.58.0+
Performance impact
The default Agent configuration for Database Monitoring is conservative, but you can adjust settings such as the collection interval and operation sampling rate to better suit your needs. For most workloads, the Agent represents less than one percent of query execution time on the database and less than one percent of CPU.
Connection strings or SRV strings
Although MongoDB connection strings or SRV strings provide many benefits such as automatic failover and load balancing, the Datadog Agent must connect directly to the MongoDB instance being monitored. If the Agent connects to a different MongoDB instance while it is running (as in the case of failover, load balancing, and so on), the Agent calculates the difference in statistics between two hosts, producing inaccurate metrics.
Data security considerations
Read about how Database Management handles sensitive information for information about what data the Agent collects from your databases and how to ensure it is secure.
Datadog Agent는 통계 및 쿼리를 수집하기 위해 MongoDB Atlas Cluster에 대한 읽기 전용 액세스 권한이 필요합니다.
커스텀 모니터링 역할 생성
MongoDB Atlas UI에서 Database Access 탭으로 이동합니다.
Custom Roles 탭에서 Add New Custom Role을 클릭합니다.
datadog 등 Custom Role Name을 입력합니다.
커스텀 역할에 다음 권한을 추가합니다.
admin 데이터베이스에 read
local 데이터베이스에 read
config 데이터베이스에 read(공유 클러스터만)
admin 데이터베이스에 clusterMonitor
모니터링하려는 사용자 생성 데이터베이스에 read 또는 모든 데이터베이스를 모니터링할 수 있는 readAnyDatabase
Add Custom Role를 클릭합니다.
모니터링 사용자와 커스텀 모니터링 역할 생성
MongoDB Atlas UI에서 Database Access 탭으로 이동합니다.
Database Users 탭에서 Add New Database User를 클릭합니다.
Authentication Method에서 Password를 선택합니다.
사용자 이름과 비밀번호를 입력합니다.
Database User Privileges에서 Custom Roles를 확대한 다음 이전 단계에서 생성한 커스텀 모니터링 역할을 선택합니다.
Add User을 클릭합니다.
모니터링 사용자의 사용자 이름 및 비밀번호를 참고하여 Agent를 구성할 수 있습니다.
비밀번호를 안전하게 저장하기
Store your password using secret management software such as Vault. You can then reference this password as ENC[<SECRET_NAME>] in your Agent configuration files: for example, ENC[datadog_user_database_password]. See Secrets Management for more information.
The examples on this page use datadog_user_database_password to refer to the name of the secret where your password is stored. It is possible to reference your password in plain text, but this is not recommended.
에이전트 설치 및 구성
MongoDB Atlas Cluster를 모니터링하려면 호스트에 Datadog Agent를 설치하고 구성하여 MongoDB Atlas Cluster에 원격으로 액세스할 수 있도록 해야 합니다. 이 호스트는 Linux 호스트, Docker 컨테이너 또는 Kubernetes 포드일 수 있습니다.
SRV 연결 호스트에서 개별 MongoDB 인스턴스 호스트 이름과 포트를 가져옵니다.
애플리케이션은 보통 SRV 연결 문자열을 통해 MongoDB Atlas에 연결합니다. 하지만 Datadog Agent는 직접 모니터링 대상인 개별 MongoDB 인스턴스에 연결해야 합니다. Agent가 실행되는 도중에 (페일오버, 로드 밸런싱 등의 경우) 각기 다른 MongoDB 인스턴스에 연결되는 경우, Agent는 두 호스트 간 통계적 차이를 계산하여 부정확한 메트릭을 생산합니다.
개별 MongoDB 인스턴스 호스트 이름 및 포트를 가져오려면 Linux의 dig 또는 Windows의 nslookup 등 네트워크 유틸리티 명령줄 도구를 사용할 수 있습니다.
복제 세트 구성원
SRV 문자열(mongodb+srv://XXXXX.XXX.mongodb.net/) 포함 비공유(복제 세트) 클러스터의 경우
To monitor a MongoDB replica set, the Agent needs to connect to all members (including the arbiter) of the replica set.
Use the following configuration block as an example to configure the Agent to connect to a replica set member:
init_config:instances:## @param hosts - required## Specify the hostname, IP address, or UNIX domain socket of## a mongod instance as listed in the replica set configuration.## If the port number is not specified, the default port 27017 is used.#- hosts:- <HOST>:<PORT>## @param username - string - optional## The username to use for authentication.#username:datadog## @param password - string - optional## The password to use for authentication.#password:"ENC[datadog_user_database_password]"## @param options - mapping - optional## Connection options. For a complete list, see:## https://docs.mongodb.com/manual/reference/connection-string/#connections-connection-options#options:authSource:admin## @param tls - boolean - optional## Set to true to connect to the MongoDB instance using TLS.#tls:true## @param dbm - boolean - optional## Set to true to enable Database Monitoring.#dbm:true## @param cluster_name - string - optional## The unique name of the cluster to which the monitored MongoDB instance belongs.## Used to group MongoDB instances in a MongoDB cluster.## cluster_name should follow Datadog tags naming conventions. See:## https://docs.datadoghq.com/developers/guide/what-best-practices-are-recommended-for-naming-metrics-and-tags/#rules-and-best-practices-for-naming-tags## Required when `dbm` is enabled.#cluster_name:<MONGO_CLUSTER_NAME>## @param reported_database_hostname - string - optional## Set the reported database hostname for the connected MongoDB instance.## This value overrides the MongoDB hostname detected by the Agent## from the MongoDB admin command serverStatus.host.#reported_database_hostname:<DATABASE_HOSTNAME_OVERRIDE>## @param additional_metrics - list of strings - optional## List of additional metrics to collect. Available options are:## - metrics.commands: Use of database commands## - tcmalloc: TCMalloc memory allocator## - top: Usage statistics for each collection## - collection: Metrics of the specified collections#additional_metrics:["metrics.commands","tcmalloc","top","collection"]## @param collections_indexes_stats - boolean - optional## Set to true to collect index statistics for the specified collections.## Requires `collections` to be set.#collections_indexes_stats:true## @param database_autodiscovery - mapping - optional## Enable database autodiscovery to automatically collect metrics from all your MongoDB databases.#database_autodiscovery:## @param enabled - boolean - required## Enable database autodiscovery.#enabled:true## @param include - list of strings - optional## List of databases to include in the autodiscovery. Use regular expressions to match multiple databases.## For example, to include all databases starting with "mydb", use "^mydb.*".## By default, include is set to ".*" and all databases are included.#include:- "^mydb.*"## @param exclude - list of strings - optional## List of databases to exclude from the autodiscovery. Use regular expressions to match multiple databases.## For example, to exclude all databases starting with "mydb", use "^mydb.*".## When the exclude list conflicts with include list, the exclude list takes precedence.#exclude:- "^mydb2.*"- "admin$"## @param max_databases - integer - optional## Maximum number of databases to collect metrics from. The default value is 100.#max_databases:100## @param refresh_interval - integer - optional## Interval in seconds to refresh the list of databases. The default value is 600 seconds.#refresh_interval:600
An example configuration for a replica set with 1 primary and 2 secondaries is as follows:
To monitor a MongoDB sharded cluster, the Agent needs to connect to the mongos router(s) and all members of the shards. If you have multiple mongos routers, you can configure the Agent to connect to them for load balancing.
Use the following configuration block as an example to configure the Agent to connect to a Mongos router:
init_config:instances:## @param hosts - required## For a sharded cluster, you need one check instance for each mongod instance in## each shard (including the configsvr shard) as well as one additional check instance## that connects to at least one mongos node.## Specify the hostname, IP address, or UNIX domain socket of the mongod or mongos instance.## If the port number is not specified, the default port 27017 is used.- hosts:- <HOST>:<PORT>## @param username - string - optional## The username to use for authentication.#username:datadog## @param password - string - optional## The password to use for authentication.#password:"ENC[datadog_user_database_password]"## @param options - mapping - optional## Connection options. For a complete list, see:## https://docs.mongodb.com/manual/reference/connection-string/#connections-connection-options#options:authSource:admin## @param tls - boolean - optional## Set to true to connect to the MongoDB instance using TLS.#tls:true## @param dbm - boolean - optional## Set to true to enable Database Monitoring.#dbm:true## @param cluster_name - string - optional## The unique name of the cluster to which the monitored MongoDB instance belongs.## Used to group MongoDB instances in a MongoDB cluster.## cluster_name should follow Datadog tags naming conventions. See:## https://docs.datadoghq.com/developers/guide/what-best-practices-are-recommended-for-naming-metrics-and-tags/#rules-and-best-practices-for-naming-tags## Required when `dbm` is enabled.#cluster_name:<MONGO_CLUSTER_NAME>## @param reported_database_hostname - string - optional## Set the reported database hostname for the connected MongoDB instance.## This value overrides the MongoDB hostname detected by## the Agent from the MongoDB admin command serverStatus.host.#reported_database_hostname:<DATABASE_HOSTNAME_OVERRIDE>## @param additional_metrics - list of strings - optional## List of additional metrics to collect. Available options are:## - metrics.commands: Use of database commands## - tcmalloc: TCMalloc memory allocator## - top: Usage statistics for each collection## - collection: Metrics of the specified collections## - jumbo_chunks: Count and percentage of jumbo chunks. Ignored on mongod instances.## - sharded_data_distribution: Distribution of data in sharded collections.#additional_metrics:["metrics.commands","tcmalloc","top","collection","jumbo_chunks","sharded_data_distribution"]## @param collections_indexes_stats - boolean - optional## Set to true to collect index statistics for the specified collections.## Requires `collections` to be set.#collections_indexes_stats:true## @param database_autodiscovery - mapping - optional## Enable database autodiscovery to automatically collect metrics from all your MongoDB databases.#database_autodiscovery:## @param enabled - boolean - required## Enable database autodiscovery.#enabled:true## @param include - list of strings - optional## List of databases to include in the autodiscovery. Use regular expressions to match multiple databases.## For example, to include all databases starting with "mydb", use "^mydb.*".## By default, include is set to ".*" and all databases are included.#include:- "^mydb.*"## @param exclude - list of strings - optional## List of databases to exclude from the autodiscovery. Use regular expressions to match multiple databases.## For example, to exclude all databases starting with "mydb", use "^mydb.*".## When the exclude list conflicts with include list, the exclude list takes precedence.#exclude:- "^mydb2.*"- "admin$"## @param max_databases - integer - optional## Maximum number of databases to collect metrics from. The default value is 100.#max_databases:100## @param refresh_interval - integer - optional## Interval in seconds to refresh the list of databases. The default value is 600 seconds.#refresh_interval:600
Refer to the Replica Set configuration for an example configuration for connecting to members in each shard and the config server.
An example configuration for a sharded cluster with 1 mongos router and 2 shards is as follows:
Place the MongoDB Agent configuration file created in the previous step in /etc/datadog-agent/conf.d/mongo.d/conf.yaml. See the sample conf file for all available configuration options.
To configure the Database Monitoring Agent running in a Docker container, set the Autodiscovery Integration Templates as Docker labels on your Agent container.
The MongoDB check is included in the Datadog Agent. No additional installation is necessary.
Note: The Agent must have read permission on the Docker socket for autodiscovery of labels to work.
Add the configuration details for the MongoDB check from the previous step in the com.datadoghq.ad.checks label. See the sample conf file for all available configuration options.
If you have a Kubernetes cluster, use the Datadog Cluster Agent for Database Monitoring.
If cluster checks are not already enabled in your Kubernetes cluster, follow the instructions to enable cluster checks. You can configure the Cluster Agent either with static files mounted in the Cluster Agent container, or by using Kubernetes service annotations.
Command line with Helm
Execute the following Helm command to install the Datadog Cluster Agent on your Kubernetes cluster. Replace the values to match your account and environment:
To configure a cluster check with a mounted configuration file, mount the configuration file in the Cluster Agent container on the path: /conf.d/mongo.yaml:
cluster_check:true# Make sure to include this flaginit_config:instances:- hosts:- <HOST>:<PORT>username:datadogpassword:"ENC[datadog_user_database_password]"options:authSource:admindbm:truecluster_name:<MONGO_CLUSTER_NAME>reported_database_hostname:<DATABASE_HOSTNAME_OVERRIDE>database_autodiscovery:enabled:trueadditional_metrics:["metrics.commands","tcmalloc","top","collection"]collections_indexes_stats:true
Configure with Kubernetes service annotations
Rather than mounting a file, you can declare the instance configuration as a Kubernetes Service. To configure this check for an Agent running on Kubernetes, create a Service in the same namespace as the Datadog Cluster Agent:
The Cluster Agent automatically registers this configuration and begins running the MongoDB integration.
To avoid exposing the datadog user’s password in plain text, use the Agent’s secret management package and declare the password using the ENC[] syntax.
MongoDB Atlas 보다 포괄적인 데이터베이스 메트릭을 수집하려면 MongoDB Atlas 통합(선택 사항)을 설치합니다.
수집한 데이터
메트릭
MongoDB 통합으로 수집한 메트릭의 포괄적인 목록을 보려면 MongoDB 통합 설명서를 참조하세요.
Slow operations
Database Monitoring for MongoDB captures slow operations from either MongoDB slow query logs or the system.profile collection. Slow operations are defined as those taking longer than the slowms threshold set in your MongoDB configuration.
With Database Profiling Enabled: When profiling is enabled at levels 1 or 2, Database Monitoring collects slow operations from the system.profile collection.
With Database Profiling Disabled: If profiling is disabled, Database Monitoring relies on MongoDB getLog command to gather slow operations from slow query logs.
Note: The getLog command retrieves the most recent 1024 mongod events. In busy databases with a high volume of slow queries or when the slow query collection interval is set to a higher interval (resulting in less frequent collection), some slow queries may not be captured.
Operation samples and explain plans
Database Monitoring for MongoDB gathers operation samples using the currentOp command. This command provides information about operations that are currently being executed on the MongoDB instance. Additionally, Database Monitoring collects explain plans for the read operation samples using the explain command, offering detailed insights into the query execution plans.
Replication state changes
Database Monitoring for MongoDB generates an event each time there is a change in the replication state within the MongoDB instance. This ensures that any changes in replication are promptly detected and reported.
Collection of schemas and indexes
Database Monitoring for MongoDB collects inferred schemas and indexes of MongoDB collections. This information is used to provide insights into the structure and organization of your collections.
When analyzing MongoDB collections, Datadog collects inferred schema information by sampling documents using the $sample aggregation stage. From this analysis, only metadata about the schema is gathered and sent to Datadog, including field names, field prevalence (how often each field appears), and their respective data types. Datadog does not collect or transmit the actual content of documents or any customer business data. This ensures that sensitive data remains protected while still providing valuable insights into the structure and organization of your collections.
