MongoDB Atlas
This Elastic integration collects logs and metrics from MongoDB Atlas instance.
Version |
0.0.3 (View all) |
Compatible Kibana version(s) |
8.13.0 or higher |
Supported Serverless project types |
Security Observability |
Subscription level |
Basic |
Level of support |
Elastic |
Overview
MongoDB Atlas, the leading multi-cloud developer data platform, offers the easiest way to run MongoDB, enabling you to work with your code's objects directly through its document-based data model, which allows for flexible schema and easy scalability.
Use the MongoDB Atlas integration to:
- Collect MongoDB Atlas mongod audit logs, mongod database logs, and process metrics for comprehensive monitoring and analysis.
- Create informative visualizations to track usage trends, measure key metrics, and derive actionable business insights.
- Set up alerts to minimize Mean Time to Detect (MTTD) and Mean Time to Resolve (MTTR) by quickly referencing relevant logs during troubleshooting.
Data streams
The MongoDB Atlas integration collects logs and metrics.
Logs help you keep a record of events that happen on your machine. The Log
data stream collected by MongoDB Atlas integration are mongod_audit
and mongod_database
.
Metrics give you insight into the statistics of the MongoDB Atlas. The Metric
data stream collected by the MongoDB Atlas integration is process
so that the user can monitor and troubleshoot the performance of the MongoDB Atlas instance.
Data streams:
-
mongod_audit
: The auditing facility allows administrators and users to track system activity for deployments with multiple users and applications. Mongod Audit logs capture events related to database operations such as insertions, updates, deletions, user authentication, etc., occurring within the mongod instances. -
mongod_database
: This datastream collects a running log of events, including entries such as incoming connections, commands run, and issues encountered. Generally, database log messages are useful for diagnosing issues, monitoring your deployment, and tuning performance. -
process
: This data stream collects host metrics per process for all the hosts of the specified group. Metrics like measurements for the host, such as CPU usage, number of I/O operations and memory are available on this data stream.
Note:
- Users can monitor and see the logs and metrics inside the ingested documents for MongoDB Atlas in the
logs-*
index pattern fromDiscover
.
Prerequisites
You can store and search your data using Elasticsearch and visualize and manage it with Kibana. We recommend using our hosted Elasticsearch Service on Elastic Cloud or self-managing the Elastic Stack on your own hardware.
Setup
To collect data from MongoDB Atlas, the following parameters from your MongoDB Atlas instance are required
- Public Key
- Private Key
- GroupId
Steps to obtain Public Key, Private Key and GroupId
- Generate programmatic API keys with
project owner
permissions by following the instructions in the Atlas documentation. Then, copy the public and private keys which function as a username and API key respectively. - From the Atlas UI, go to Project Settings > Access Manager > API Keys and then click on Invite To Project to add the API key created above.
- Add specific role to API keys, under Project Settings > Access Manager > API Keys. This step is important to make sure that these API keys have the right permissions to access the data without running into any issues. The specific role for each datastream is defined under data stream reference section.
- Enable Database Auditing for the Atlas project you want to monitor logs. You can follow the instructions provided in this Atlas document.
- You can find your Project ID (Group ID) in the Atlas UI. To do this, navigate to your project, click on Settings, and copy the Project ID (Group ID). You can also programmatically find it using the Atlas Admin API or Atlas CLI as described in this Atlas document.
Important terms of MongoDB Atlas API
- Granularity: Duration that specifies the interval at which Atlas reports the metrics.
- Period: Duration over which Atlas reports the metrics.
Note: Both of above attributes can be set by using period
in configuration parameters.
Steps to enable Integration in Elastic
- In Kibana go to Management > Integrations
- In "Search for integrations" search bar, type MongoDB Atlas
- Click on the "MongoDB Atlas" integration from the search results.
- To add the integration, click on the "Add MongoDB Atlas" button.
- Enter all the necessary configuration parameters, including Public Key, Private Key, and GroupId.
- Finally, save the integration.
Note:
- The
mongod_audit
andmongod_database
data streams gather historical data spanning the previous 30 minutes. - We recommend setting an interval of five minutes or higher for collecting mongod audit and database logs, as MongoDB Atlas refreshes logs from the cluster's backend infrastructure at five minutes intervals as described in this Atlas document.
- The logs collection from MongoDB Atlas does not support M0 free clusters, M2/M5 shared clusters, or serverless instances.
- Mongod: Mongod is the primary daemon method for the MongoDB system. It helps in handling the data requests, managing the data access, performing background management operations, and other core database operations.
Troubleshooting
If you encounter an error while ingesting data, it might be due to the data collected over a long time span. Generating a response in such cases may take longer and might cause a request timeout if the HTTP Client Timeout
parameter is set to a small duration. To avoid this error, it is recommended to adjust the HTTP Client Timeout
and Interval
parameters based on the duration of data collection.
{
"error": {
"message": "failed eval: net/http: request canceled (Client.Timeout or context cancellation while reading body)"
}
}
Logs reference
Mongod Audit
This is the mongod_audit
data stream. This data stream allows administrators and users to track system activity for deployments with multiple users and applications. To collect audit logs, the requesting API Key must have the Project Data Access Read Only
or higher role.
An example event for mongod_audit
looks as following:
{
"@timestamp": "2023-04-01T12:00:00.000Z",
"agent": {
"ephemeral_id": "82e5a7ce-c7ad-436f-90f2-a1cefbe22333",
"id": "498e0c10-4447-4a56-90f0-ba02c44a01c2",
"name": "docker-fleet-agent",
"type": "filebeat",
"version": "8.13.0"
},
"data_stream": {
"dataset": "mongodb_atlas.mongod_audit",
"namespace": "ep",
"type": "logs"
},
"ecs": {
"version": "8.11.0"
},
"elastic_agent": {
"id": "498e0c10-4447-4a56-90f0-ba02c44a01c2",
"snapshot": false,
"version": "8.13.0"
},
"event": {
"action": "authenticate",
"agent_id_status": "verified",
"category": [
"network",
"authentication"
],
"dataset": "mongodb_atlas.mongod_audit",
"ingested": "2024-04-09T06:13:37Z",
"kind": "event",
"module": "mongodb_atlas",
"type": [
"access",
"info"
]
},
"host": {
"architecture": "x86_64",
"containerized": true,
"hostname": "docker-fleet-agent",
"id": "8259e024976a406e8a54cdbffeb84fec",
"ip": [
"192.168.245.7"
],
"mac": [
"02-42-C0-A8-F5-07"
],
"name": "docker-fleet-agent",
"os": {
"codename": "focal",
"family": "debian",
"kernel": "3.10.0-1160.92.1.el7.x86_64",
"name": "Ubuntu",
"platform": "ubuntu",
"type": "linux",
"version": "20.04.6 LTS (Focal Fossa)"
}
},
"input": {
"type": "cel"
},
"mongodb_atlas": {
"mongod_audit": {
"hostname": "hostname1",
"local": {
"ip": "127.0.0.1",
"port": 27017
},
"remote": {
"ip": "192.168.1.100",
"port": 54320
},
"result": "Success",
"user": {
"names": [
{
"db": "admin",
"user": "auditUser"
}
],
"roles": [
{
"db": "admin",
"role": "dbAdmin"
}
]
},
"uuid": {
"binary": "some-unique-identifier",
"type": "04"
}
}
},
"related": {
"ip": [
"127.0.0.1",
"192.168.1.100"
]
},
"tags": [
"mongodb_atlas-mongod_audit"
]
}
Exported fields
Field | Description | Type |
---|---|---|
@timestamp | Event timestamp. | date |
data_stream.dataset | Data stream dataset. | constant_keyword |
data_stream.namespace | Data stream namespace. | constant_keyword |
data_stream.type | Data stream type. | constant_keyword |
input.type | Type of Filebeat input. | keyword |
mongodb_atlas.mongod_audit.hostname | Hostname that stores the log files that you want to download. | keyword |
mongodb_atlas.mongod_audit.local.ip | IP address of the running instance. | ip |
mongodb_atlas.mongod_audit.local.is_system_user | True if the event is caused by a system user, false otherwise. | boolean |
mongodb_atlas.mongod_audit.local.port | Port number of the running instance. | long |
mongodb_atlas.mongod_audit.local.unix | Unix that contains the MongoDB socket file path if the client connects through a Unix domain socket. | keyword |
mongodb_atlas.mongod_audit.param | Specific details for the event. | object |
mongodb_atlas.mongod_audit.remote.ip | IP address of the incoming connection associated with the event. | ip |
mongodb_atlas.mongod_audit.remote.is_system_user | True if the event is caused by a system user, false otherwise. | boolean |
mongodb_atlas.mongod_audit.remote.port | Port number of the incoming connection associated with the event. | long |
mongodb_atlas.mongod_audit.remote.unix | Unix that contains the MongoDB socket file path if the client connects through a Unix domain socket. | keyword |
mongodb_atlas.mongod_audit.result | Error code. | keyword |
mongodb_atlas.mongod_audit.user.names | Array of user identification documents. | object |
mongodb_atlas.mongod_audit.user.roles | Array of documents that specify the roles granted to the user. | object |
mongodb_atlas.mongod_audit.uuid.binary | Document that contains a universally unique identifier (UUID) for the audit message. | keyword |
mongodb_atlas.mongod_audit.uuid.type | The $type field specifies the BSON subtype for the $binary field. | keyword |
Mongod Database
This is the mongod_database
data stream. This datastream collects a running log of events, including entries such as incoming connections, commands run, monitoring deployment, tuning performance, and issues encountered. To collect database logs, the requesting API Key must have the Project Data Access Read Only
or higher role.
An example event for mongod_database
looks as following:
{
"@timestamp": "2024-02-18T14:45:23.512Z",
"agent": {
"ephemeral_id": "dbad1b64-5ae2-467e-a76a-7d31d2bbc35a",
"id": "35b61223-ca83-481f-a4aa-ab5983a75ba8",
"name": "docker-fleet-agent",
"type": "filebeat",
"version": "8.13.0"
},
"data_stream": {
"dataset": "mongodb_atlas.mongod_database",
"namespace": "ep",
"type": "logs"
},
"ecs": {
"version": "8.11.0"
},
"elastic_agent": {
"id": "35b61223-ca83-481f-a4aa-ab5983a75ba8",
"snapshot": false,
"version": "8.13.0"
},
"event": {
"agent_id_status": "verified",
"category": [
"network",
"database"
],
"dataset": "mongodb_atlas.mongod_database",
"ingested": "2024-04-05T10:24:59Z",
"kind": "event",
"module": "mongodb_atlas",
"type": [
"access",
"info"
]
},
"host": {
"architecture": "x86_64",
"containerized": true,
"hostname": "docker-fleet-agent",
"id": "8259e024976a406e8a54cdbffeb84fec",
"ip": [
"192.168.255.7"
],
"mac": [
"02-42-C0-A8-FF-07"
],
"name": "docker-fleet-agent",
"os": {
"codename": "focal",
"family": "debian",
"kernel": "3.10.0-1160.92.1.el7.x86_64",
"name": "Ubuntu",
"platform": "ubuntu",
"type": "linux",
"version": "20.04.6 LTS (Focal Fossa)"
}
},
"input": {
"type": "cel"
},
"log": {
"level": "informational"
},
"mongodb_atlas": {
"mongod_database": {
"component": "NETWORK",
"hostname": "hostname1",
"id": 67890,
"message": "Client connection accepted",
"tags": [
"connection"
],
"thread": {
"name": "conn123"
}
}
},
"tags": [
"mongodb_atlas-mongod_database"
]
}
Exported fields
Field | Description | Type |
---|---|---|
@timestamp | Event timestamp. | date |
data_stream.dataset | Data stream dataset. | constant_keyword |
data_stream.namespace | Data stream namespace. | constant_keyword |
data_stream.type | Data stream type. | constant_keyword |
input.type | Type of Filebeat input. | keyword |
mongodb_atlas.mongod_database.attributes | One or more key-value pairs for additional log attributes. If a log message does not include any additional attributes, the attr object is omitted. | object |
mongodb_atlas.mongod_database.component | The component field indicates the category to which a logged event belongs, such as NETWORK or COMMAND. | keyword |
mongodb_atlas.mongod_database.hostname | A human-readable label that identifies the host that stores the log files you want to download. | keyword |
mongodb_atlas.mongod_database.id | The unique identifier for the log statement. | long |
mongodb_atlas.mongod_database.message | The log output message passed from the server or driver. If necessary, the message is escaped according to the JSON specification. | match_only_text |
mongodb_atlas.mongod_database.size | The original size of a log entry if it has been truncated. Only included if the log entry contains at least one truncated attr attribute. | object |
mongodb_atlas.mongod_database.tags | Strings representing any tags applicable to the log statement, for example, ["startupWarnings"]. | keyword |
mongodb_atlas.mongod_database.thread.name | The name of the thread that caused the log statement. | keyword |
mongodb_atlas.mongod_database.truncated | Information about log message truncation, if applicable. Only included if the log entry contains at least one truncated attr attribute. | object |
Metrics reference
Process
This data stream collects host metrics per process for all the hosts of the specified group. Metrics like measurements for the host, such as CPU usage, number of I/O operations and memory are available on this data stream. To collect process metrics, the requesting API Key must have the Project Read Only
role.
An example event for process
looks as following:
{
"@timestamp": "2024-04-11T12:42:53.267Z",
"agent": {
"ephemeral_id": "c8ebb866-6d72-471b-9083-6d386219bf61",
"id": "926ca6d4-5487-4a8b-b88b-34f188fe8cfb",
"name": "docker-fleet-agent",
"type": "filebeat",
"version": "8.13.0"
},
"data_stream": {
"dataset": "mongodb_atlas.process",
"namespace": "ep",
"type": "logs"
},
"ecs": {
"version": "8.11.0"
},
"elastic_agent": {
"id": "926ca6d4-5487-4a8b-b88b-34f188fe8cfb",
"snapshot": false,
"version": "8.13.0"
},
"event": {
"agent_id_status": "verified",
"category": [
"process"
],
"dataset": "mongodb_atlas.process",
"ingested": "2024-04-11T12:43:05Z",
"kind": "event",
"module": "mongodb_atlas",
"type": [
"info"
]
},
"host": {
"architecture": "x86_64",
"containerized": true,
"hostname": "docker-fleet-agent",
"id": "8259e024976a406e8a54cdbffeb84fec",
"ip": [
"192.168.253.4"
],
"mac": [
"02-42-C0-A8-FD-04"
],
"name": "docker-fleet-agent",
"os": {
"codename": "focal",
"family": "debian",
"kernel": "3.10.0-1160.92.1.el7.x86_64",
"name": "Ubuntu",
"platform": "ubuntu",
"type": "linux",
"version": "20.04.6 LTS (Focal Fossa)"
}
},
"input": {
"type": "cel"
},
"mongodb_atlas": {
"group_id": "mongodb-group1",
"host_id": "hostname1",
"process": {
"assert": {
"regular": 0
}
},
"process_id": "hostname1"
},
"tags": [
"mongodb_atlas-process"
]
}
Exported fields
Field | Description | Type | Unit | Metric Type |
---|---|---|---|---|
@timestamp | Event timestamp. | date | ||
data_stream.dataset | Data stream dataset. | constant_keyword | ||
data_stream.namespace | Data stream namespace. | constant_keyword | ||
data_stream.type | Data stream type. | constant_keyword | ||
input.type | Type of Filebeat input. | keyword | ||
mongodb_atlas.group_id | Identifier for the project of the event. | keyword | ||
mongodb_atlas.host_id | Unique identifier of the host for the MongoDB process. | keyword | ||
mongodb_atlas.process.assert.msg | The average rate of message asserts per second over the selected sample period. | double | gauge | |
mongodb_atlas.process.assert.regular | The average rate of regular asserts raised per second over the selected sample period. | double | gauge | |
mongodb_atlas.process.assert.user | The average rate of user asserts per second over the selected sample period. | double | gauge | |
mongodb_atlas.process.assert.warning | The average rate of warnings per second over the selected sample period. | double | gauge | |
mongodb_atlas.process.background_flush.avg | Amount of data flushed in the background. | double | gauge | |
mongodb_atlas.process.cache.dirty.bytes | Write - Amount of bytes in the WiredTiger storage engine cache. | double | byte | gauge |
mongodb_atlas.process.cache.read.bytes | Read - Amount of bytes in the WiredTiger storage engine cache. | double | byte | gauge |
mongodb_atlas.process.cache.used.total.bytes | The total bytes cached in memory for serving reads and writes. | double | byte | gauge |
mongodb_atlas.process.cache.write.bytes | The maximum disk read latency value over the period specified by the metric granularity. | double | byte | gauge |
mongodb_atlas.process.connections | Displays the total number of active connections to the database deployment. Monitor connections to determine whether the current connection limits are sufficient. | double | gauge | |
mongodb_atlas.process.cpu.children.kernel.max.pct | The maximum amount of CPU time spent by child processes in kernel space. | double | percent | counter |
mongodb_atlas.process.cpu.children.kernel.pct | CPU children kernel space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cpu.children.user.max.pct | The maximum amount of CPU time spent by child processes in user space. | double | percent | counter |
mongodb_atlas.process.cpu.children.user.pct | CPU children user space for mongodb. | double | percent | gauge |
mongodb_atlas.process.cpu.kernel.max.pct | The maximum amount of CPU time spent by the MongoDB process itself in kernel space, handling system calls and hardware interrupts. | double | percent | counter |
mongodb_atlas.process.cpu.kernel.pct | CPU kernel space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cpu.normalized.children.kernel.max.pct | Max children kernel CPU usage scaled to a range of 0% to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.cpu.normalized.children.kernel.pct | NORMALIZED CPU children kernel space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cpu.normalized.children.user.max.pct | Max children user CPU usage scaled to a range of 0% to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.cpu.normalized.children.user.pct | NORMALIZED CPU children user space for mongodb. | double | percent | gauge |
mongodb_atlas.process.cpu.normalized.kernel.max.pct | Max kernel CPU usage scaled to a range of 0% to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.cpu.normalized.kernel.pct | NORMALIZED CPU kernel space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cpu.normalized.user.max.pct | Max user CPU usage scaled to a range of 0% to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.cpu.normalized.user.pct | NORMALIZED CPU user space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cpu.user.max.pct | The maximum amount of CPU time spent by the MongoDB process itself in user space, executing application code and processing data. | double | percent | counter |
mongodb_atlas.process.cpu.user.pct | CPU user space for mongodb processes. | double | percent | gauge |
mongodb_atlas.process.cursor.open.total | The number of cursors that the server maintains for clients. | long | gauge | |
mongodb_atlas.process.cursor.timed_out.total | The average rate of cursors that have timed out per second over the selected sample period. | double | gauge | |
mongodb_atlas.process.database.size.total.bytes | The amount of storage space in bytes that your stored data uses. | long | byte | gauge |
mongodb_atlas.process.database.storage.total.bytes | Sum total of the compressed on-disk storage space allocated for document storage across all databases. | long | byte | gauge |
mongodb_atlas.process.document.deleted | Displays the documents deleted per second. | double | gauge | |
mongodb_atlas.process.document.inserted | Displays the documents inserted per second. | double | gauge | |
mongodb_atlas.process.document.returned | Displays the documents returned per second. | double | gauge | |
mongodb_atlas.process.document.updated | Displays the documents updated per second. | double | gauge | |
mongodb_atlas.process.fts.cpu.kernel.pct | The amount of CPU time spent by the Full-Text search process in kernel space. | double | percent | gauge |
mongodb_atlas.process.fts.cpu.normalized.kernel.pct | Percentage of time that the CPU spent servicing the operating system calls for the search process. | double | percent | gauge |
mongodb_atlas.process.fts.cpu.normalized.user.pct | Percentage of time that the CPU spent servicing user calls for the search process. | double | percent | gauge |
mongodb_atlas.process.fts.cpu.user.pct | The amount of CPU time spent by the Full-Text search process in user space. | double | percent | gauge |
mongodb_atlas.process.fts.disk.utilization.total.bytes | Total bytes of disk space that search processes use. | long | byte | gauge |
mongodb_atlas.process.fts.memory.mapped.total.bytes | Total bytes of mapped memory that search processes occupy. | long | byte | gauge |
mongodb_atlas.process.fts.memory.resident.total.bytes | Total bytes of resident memory that search processes occupy. | long | byte | gauge |
mongodb_atlas.process.fts.memory.virtual.total.bytes | Total bytes of virtual memory that search processes occupy. | long | byte | gauge |
mongodb_atlas.process.global.access.not_in_memory | The number of accesses to data that are not currently stored in memory, requiring disk access. | long | gauge | |
mongodb_atlas.process.global.lock.current_queue.reader.count | The number of operations that are currently queued and waiting for the read lock. | long | gauge | |
mongodb_atlas.process.global.lock.current_queue.total | The total number of operations queued waiting for the lock (readers + writers) | long | gauge | |
mongodb_atlas.process.global.lock.current_queue.writer.count | The number of operations that are currently queued and waiting for the write lock. | long | gauge | |
mongodb_atlas.process.global.page_fault.exception_thrown | The number of exceptions thrown due to page faults. | long | gauge | |
mongodb_atlas.process.host.page_faults | Measurements on page faults related to the host. | double | gauge | |
mongodb_atlas.process.index.btree.access.count | Number of index btree ACCESSES. | long | gauge | |
mongodb_atlas.process.index.btree.hits.count | Number of index btree HITS. | long | gauge | |
mongodb_atlas.process.index.btree.miss.count | Number of index btree MISSES. | long | gauge | |
mongodb_atlas.process.index.btree.miss_ratio.count | Index btree miss ratio. | double | gauge | |
mongodb_atlas.process.journaling.commits.write_lock | Number of journaling COMMIT operations. | long | gauge | |
mongodb_atlas.process.journaling.mb | Average amount of data in megabytes Cloud Manager writes to the recovery log per second meets your specified threshold. | double | gauge | |
mongodb_atlas.process.journaling.write.data_files.mb | The maximum size in megabytes (MB) of data files written by the journaling process. | double | gauge | |
mongodb_atlas.process.memory.computed.mb | Amount of COMPUTED process memory in megabytes. | double | gauge | |
mongodb_atlas.process.memory.mapped.mb | Amount of MAPPED process memory in megabytes. | double | gauge | |
mongodb_atlas.process.memory.resident.mb | Amount of RESIDENT process memory in megabytes. | double | gauge | |
mongodb_atlas.process.memory.virtual.mb | Amount of VIRTUAL process memory in megabytes. | double | gauge | |
mongodb_atlas.process.network.in | Process incoming network throughput in bytes per second. | double | gauge | |
mongodb_atlas.process.network.out | Process outgoing network throughput in bytes per second. | double | gauge | |
mongodb_atlas.process.network.request.total | The total number of distinct requests that the process has received. | double | counter | |
mongodb_atlas.process.opcounter.cmd | Database operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.opcounter.delete | Database DELETE operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.opcounter.getmore | Database GETMORE operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.opcounter.insert | Database INSERT operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.opcounter.query | Database QUERY operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.opcounter.repl.cmd | Database operations rate on secondaries. | double | gauge | |
mongodb_atlas.process.opcounter.repl.delete | Database DELETE operations rate on secondaries. | double | gauge | |
mongodb_atlas.process.opcounter.repl.insert | Database INSERT operations rate on secondaries. | double | gauge | |
mongodb_atlas.process.opcounter.repl.update | Database UPDATE operations rate on secondaries. | double | gauge | |
mongodb_atlas.process.opcounter.update | Database UPDATE operations rate on a process since the process last started. | double | gauge | |
mongodb_atlas.process.operation.execution.time.cmd.avg.ms | Average execution time in milliseconds per command operation over the selected sample period. | double | ms | gauge |
mongodb_atlas.process.operation.execution.time.read.avg.ms | Average execution time in milliseconds per read operation over the selected sample period. | double | ms | gauge |
mongodb_atlas.process.operation.execution.time.write.avg.ms | Average execution time in milliseconds per write operation over the selected sample period. | double | ms | gauge |
mongodb_atlas.process.operation.scan_and_order | The total number of queries that return sorted data that cannot perform the sort operation using an index. | double | gauge | |
mongodb_atlas.process.oplog.master.lag.time_diff.s | Lag or delay in replication between the primary node (oplog master) and its secondary nodes. | double | s | gauge |
mongodb_atlas.process.oplog.master.time.s | The replication oplog window. The approximate time available in the primary's replication oplog. If a secondary is behind real-time by more than this amount, it cannot catch up and will require a full resync. | double | s | gauge |
mongodb_atlas.process.oplog.rate.gb_per_hour | The rate of change in the size of the oplog in gigabytes per hour. | double | gauge | |
mongodb_atlas.process.oplog.repl_lag.s | The amount of time, typically in seconds, it takes for changes recorded in the oplog on the primary node to be replicated and applied to the secondary node. | double | s | gauge |
mongodb_atlas.process.oplog.slave.lag.master.time.s | The difference in time, typically in seconds, between the oplog time on the secondary (slave) node and the time of the latest operation in the oplog on the primary (master) node. | double | s | gauge |
mongodb_atlas.process.query.executor.scanned | Average rate per second to scan index items during queries and query-plan evaluations. | double | gauge | |
mongodb_atlas.process.query.executor.scanned_objects | Average rate of documents scanned per second during queries and query-plan evaluations. | double | gauge | |
mongodb_atlas.process.query.targeting.scanned_objects_per_returned | Ratio of the number of documents scanned to the number of documents returned. | double | gauge | |
mongodb_atlas.process.query.targeting.scanned_per_returned | Ratio of the number of index items scanned to the number of documents returned. | double | gauge | |
mongodb_atlas.process.restart.in_last_hour | Number of times the host restarted within the previous hour. | double | gauge | |
mongodb_atlas.process.swap.usage.free.max.kb | Max amount of swap space free. | double | counter | |
mongodb_atlas.process.swap.usage.total.free | Total amount of swap space free. | double | gauge | |
mongodb_atlas.process.swap.usage.total.used | Total amount of swap space in use. | double | gauge | |
mongodb_atlas.process.swap.usage.used.max.kb | Max amount of swap space in use. | double | counter | |
mongodb_atlas.process.system.cpu.guest.max.pct | Max amount of CPU time spent running a virtual CPU for guest operating systems. | double | percent | counter |
mongodb_atlas.process.system.cpu.guest.pct | Tracks CPU time consumed by guest operating systems, like virtual machines, on the host system. | double | percent | gauge |
mongodb_atlas.process.system.cpu.iowait.max.pct | Max amount of CPU time spent waiting for I/O operations to complete. | double | percent | counter |
mongodb_atlas.process.system.cpu.iowait.pct | The CPU time spent waiting for I/O operations to complete. | double | percent | gauge |
mongodb_atlas.process.system.cpu.irq.max.pct | Max amount of CPU time spent servicing hardware interrupts. | double | percent | counter |
mongodb_atlas.process.system.cpu.irq.pct | The portion of CPU time spent servicing hardware interrupts. | double | percent | gauge |
mongodb_atlas.process.system.cpu.kernel.max.pct | Max amount of CPU time spent in kernel space, executing system calls and handling hardware interrupts. | double | percent | counter |
mongodb_atlas.process.system.cpu.kernel.pct | The portion of CPU time spent executing kernel space processes and handling system calls. | double | percent | gauge |
mongodb_atlas.process.system.cpu.nice.pct | The portion of CPU time allocated to processes with a 'nice' priority level. | double | percent | gauge |
mongodb_atlas.process.system.cpu.softirq.max.pct | Max amount of CPU time spent servicing soft interrupts, which are interrupts triggered by software. | double | percent | counter |
mongodb_atlas.process.system.cpu.softirq.pct | The CPU time spent handling software-generated interrupts. | double | percent | gauge |
mongodb_atlas.process.system.cpu.steal.max.pct | Max amount of CPU time 'stolen' by the hypervisor for other virtual machines running on the same physical host. | double | percent | counter |
mongodb_atlas.process.system.cpu.steal.pct | The amount of CPU time 'stolen' by the hypervisor for other virtual machines running on the same physical host. | double | percent | gauge |
mongodb_atlas.process.system.cpu.user.max.pct | Max amount of CPU time spent in user space, executing user-level processes and applications. | double | percent | counter |
mongodb_atlas.process.system.cpu.user.pct | The portion of CPU time spent executing user space processes and running applications. | double | percent | gauge |
mongodb_atlas.process.system.memory.available.kb | Physical memory available in kilobytes. | double | gauge | |
mongodb_atlas.process.system.memory.available.max.kb | Max Physical memory available in kilobytes. | double | counter | |
mongodb_atlas.process.system.memory.free.kb | Physical memory free in kilobytes. | double | gauge | |
mongodb_atlas.process.system.memory.free.max.kb | Max Physical memory free in kilobytes. | double | counter | |
mongodb_atlas.process.system.memory.used.kb | Physical memory used in kilobytes. | double | gauge | |
mongodb_atlas.process.system.memory.used.max.kb | Max Physical memory used in kilobytes. | double | counter | |
mongodb_atlas.process.system.network.in | Incoming network throughput in bytes per second. | double | gauge | |
mongodb_atlas.process.system.network.max.in | Max Incoming network throughput in bytes per second. | double | counter | |
mongodb_atlas.process.system.network.max.out | Max Outgoing network throughput in bytes per second. | double | counter | |
mongodb_atlas.process.system.network.out | Outgoing network throughput in bytes per second. | double | gauge | |
mongodb_atlas.process.system.normalized.cpu.guest.max.pct | Max Guest CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.guest.pct | Guest CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.iowait.max.pct | Max CPU usage of processes spent waiting for IO operations to complete, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.iowait.pct | CPU usage of processes spent waiting for IO operations to complete, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.irq.max.pct | Max CPU usage of hardware interrupts, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.irq.pct | NORMALIZED irq CPU utilization across various processes for a server | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.kernel.max.pct | Max kernel CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.kernel.pct | Kernel CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.nice.max.pct | Max CPU usage of processes with a positive nice value, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.nice.pct | NORMALIZED nice CPU utilization across various processes for a server | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.softirq.max.pct | Max CPU usage of software interrupts, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.softirq.pct | NORMALIZED softirq CPU utilization across various processes for a server | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.steal.max.pct | Max The percentage of time the CPU had something runnable, but the hypervisor chose to run something else, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.steal.pct | The percentage of time the CPU had something runnable, but the hypervisor chose to run something else, scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.steal.user.max.pct | Max user CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.steal.user.pct | NORMALIZED CPU user space utilization across various processes for a server | double | percent | gauge |
mongodb_atlas.process.system.normalized.cpu.user.max.pct | Max user CPU usage of processes on the host scaled to a range of 0 to 100%, is obtained by dividing the usage value by the total number of CPU cores. | double | percent | counter |
mongodb_atlas.process.system.normalized.cpu.user.pct | The portion of CPU time spent executing user space processes and running applications. | double | percent | gauge |
mongodb_atlas.process.ticket.available.read.count | The number of read tickets available to the WiredTiger storage engine. | long | gauge | |
mongodb_atlas.process.ticket.available.write.count | The number of write tickets available to the WiredTiger storage engine. | long | gauge | |
mongodb_atlas.process_id | Combination of hostname and Internet Assigned Numbers Authority (IANA) port that serves the MongoDB process. | keyword |
Changelog
Version | Details | Kibana version(s) |
---|---|---|
0.0.3 | Enhancement View pull request | — |
0.0.2 | Enhancement View pull request | — |
0.0.1 | Enhancement View pull request | — |