MongoDB

Integrating MongoDB Metrics and Logs to KubeSense

KubeSense ingests MongoDB telemetry through the OpenTelemetry (OTEL) Collector, letting you visualize database performance and logs alongside the rest of your infrastructure.

Prerequisites

Before you begin, ensure you have:

A MongoDB deployment running version 4.4 or newer (standalone, replica set, or sharded cluster)
KubeSense installed with an OTEL Collector that can reach the MongoDB instances
A MongoDB monitoring user with clusterMonitor permissions
File access to the MongoDB logs (structured logging recommended) if you plan to collect logs
Access to modify the KubeSense Helm values file

Collecting MongoDB Metrics

Step 1: Update KubeSense Helm Values File

Add the MongoDB receiver configuration to your KubeSense Helm values file inside the kubesensor section:

otel-collector:
  config:
    receivers:
      mongodb:
        hosts:
          - endpoint: <MONGODB_HOST>:27017
        # Provide credentials for the monitoring user if authentication is enabled
        username: <username>
        password: <password>
        collection_interval: 60s
        tls:
          insecure: true
          insecure_skip_verify: true
        metrics:
          mongodb.lock.acquire.count:
            enabled: true
          mongodb.lock.acquire.time:
            enabled: true
          mongodb.lock.deadlock.count:
            enabled: true
          mongodb.operation.latency.time:
            enabled: true
    appendToPipelines:
      metrics:
        receivers:
        - mongodb

Note: If you are monitoring a replica set or sharded deployment, list each endpoint under hosts. For TLS-enabled clusters, replace the tls block with your CA and client certificate paths.

Step 2: Apply the Configuration

Upgrade your KubeSense deployment to load the updated values:

helm upgrade kubesense kubesense/kubesense \
  -f values.yaml \
  --namespace kubesense

Collecting MongoDB Logs

Step 1: Enable Log Collection in Helm Values

Configure the filelog receiver to parse structured MongoDB logs:

otel-collector:
  config:
    receivers:
      filelog/mongodb:
        include: ["/var/log/mongodb/mongodb.log"]
        operators:
          - type: json_parser
            if: body matches '^{.*}$'
            parse_from: body
            parse_to: attributes
            timestamp:
              parse_from: attributes.t.$$date
              layout: '2006-01-02T15:04:05.000-07:00'
              layout_type: gotime
            severity:
              parse_from: attributes.s
              overwrite_text: true
              mapping:
                debug: [D1, D2, D3, D4, D5]
                info: I
                warn: W
                error: E
                fatal: F
          - type: flatten
            if: attributes.attr != nil
            field: attributes.attr
          - type: move
            if: attributes.msg != nil
            from: attributes.msg
            to: body
          - type: move
            if: attributes.c != nil
            from: attributes.c
            to: attributes.component
          - type: move
            if: attributes.id != nil
            from: attributes.id
            to: attributes.mongo_log_id
          - type: remove
            if: attributes.t != nil
            field: attributes.t
          - type: remove
            if: attributes.s != nil
            field: attributes.s
          - type: add
            field: attributes.source
            value: mongodb
    processors:
      batch/mongodb:
        send_batch_size: 10000
        send_batch_max_size: 11000
        timeout: 10s
    service:
      pipelines:
        logs/mongodb:
          receivers:
          - filelog/mongodb
          processors:
          - batch
          - resource
          exporters:
          - otlphttp/kubesense_logs

Note: Update the include path to match your MongoDB log file location. Structured logging must be enabled (systemLog.component.control: { verbosity: 1 }) for the parser to work correctly.

Step 2: Apply the Configuration

Redeploy KubeSense with the updated values:

helm upgrade kubesense kubesense/kubesense \
  -f values.yaml \
  --namespace kubesense

MongoDB Metrics Collected

KubeSense captures the following key MongoDB metrics:

Metric Name	Type	Unit	Description
mongodb_connection_count	sum	number	Total number of client connections
mongodb_operation_count	sum	number	Total operations executed
mongodb_operation_latency_time	gauge	microseconds	Operation latency
mongodb_lock_acquire_count	sum	number	Number of lock acquisitions
mongodb_lock_acquire_wait_count	sum	number	Lock waits due to contention
mongodb_lock_acquire_time	sum	microseconds	Time spent waiting for locks
mongodb_memory_usage	sum	bytes	MongoDB process memory usage
mongodb_data_size	sum	bytes	Data size across databases
mongodb_index_size	sum	bytes	Total index size
mongodb_uptime	sum	milliseconds	MongoDB server uptime

MongoDB Log Attributes

Name	Path	Type	Description
Timestamp	timestamp	timestamp	Log entry timestamp
Severity Text	severity_text	string	Parsed severity level
Severity Number	severity_number	number	Numeric severity
Component	attributes.component	string	MongoDB subsystem
Message	body	string	Log message content
Log ID	attributes.mongo_log_id	string	MongoDB log identifier

Visualization in KubeSense

Once the pipelines are active, MongoDB metrics and logs appear in KubeSense dashboards and the log explorer, enabling you to:

Track connections, latency, and lock contention
Monitor storage growth and memory usage
Investigate operational and audit events from structured logs
Build alerts for key MongoDB health indicators

Configuration Parameters

Parameter	Type	Required	Description
`hosts[].endpoint`	string	Yes	MongoDB host and port
`username`	string	Conditional	Monitoring user name when auth is enabled
`password`	string	Conditional	Monitoring user password when auth is enabled
`collection_interval`	string	No	Metrics polling interval (`60s` default)
`tls.*`	string/boolean	No	TLS connection options
`include` (logs)	array	Yes	Log file paths accessible to the collector

Troubleshooting

Issue: Metrics not collected

Verify the OTEL collector can reach each MongoDB host
Ensure the monitoring user has clusterMonitor privileges
Check OTEL collector logs for authentication or TLS errors

Issue: Logs missing or unparsed

Confirm MongoDB structured logging is enabled
Verify the log file path and permissions inside the collector pod
Review the OTEL collector logs for filelog parser errors

Issue: High lock wait counts

Investigate workload contention and slow operations in MongoDB
Consider building alerts on mongodb_lock_acquire_wait_count to detect bottlenecks

Best Practices

Isolate MongoDB metrics pipelines if monitoring multiple clusters
Rotate MongoDB log files and ensure the collector can follow rotations
Use resource attributes to label clusters (kubesense.cluster, kubesense.env_type)
Regularly review collector logs for API or parser warnings

Conclusion

Integrating MongoDB with KubeSense provides unified visibility into database performance, capacity, and operational events, helping teams troubleshoot faster and maintain healthy clusters.

MongoDB

ON THIS PAGE