MongoDB Integration

Monitor MongoDB operations, query performance, storage growth, index usage and replica set health from a single Watchlog agent.

The agent runs mongosh once per collection cycle and derives everything from that single round-trip — server status, per-database and per-collection storage, index usage, replication, and slow queries when you have profiling enabled.

Supported Platforms

• Ubuntu
• Windows
• Source Code
• Docker Container
• Kubernetes

Requirements

Required Privileges

The built-in clusterMonitor role covers almost everything. Add read on the databases you want storage and index detail for:

use admin
db.createUser({
  user: "watchlog_monitor",
  pwd: "your_password",
  roles: [
    { role: "clusterMonitor", db: "admin" },
    { role: "readAnyDatabase", db: "admin" }
  ]
})

If you would rather not grant readAnyDatabase, grant read per database instead — the databases you do not grant simply do not appear.

What each privilege buys:

What Watchlog Collects

Server Version, uptime, current and available connections, resident and virtual memory, operation counters (insert, query, update, delete, command), network in/out and request count, and per-interval latency averages for reads, writes, commands and transactions.

WiredTiger Cache used against configured maximum, dirty cache percentage, and read/write ticket availability — the signal that the storage engine is the bottleneck rather than the query.

Databases and collections Per database: data size, storage size, index size, object count, average object size, collection and index counts. Per collection: documents, size, storage, index size, index count, and capped status.

Indexes Per index: size and accesses since the server started, with unused-index candidates flagged. Primary key and unique indexes are never proposed for removal — they enforce correctness, not performance.

Replication Set name, member states and health, elected primary, per-member replication lag, and election counts.

Slow queries (when profiling is enabled — see below) Namespace, operation, duration, documents examined against returned, plan summary, and the normalized query shape.

Health Score and Insights

Watchlog computes a 0–100 score server-side, with a written reason attached to every deduction:

Insights add the interpretation: collections scanning far more documents than they return, indexes that recorded zero accesses, databases growing unusually fast, and replica members falling behind.

Slow Queries

MongoDB writes slow operations to system.profile, but only when the database profiler is enabled — and it is off by default.

// Profile operations slower than 100ms on this database
use myapp
db.setProfilingLevel(1, { slowms: 100 })

Then enable collection in the agent config:

"slowQuery": {
  "enabled": true,
  "threshold": 100,
  "maxPerScrape": 100
}

The agent tracks a high-water mark per instance, so the same profile document is never shipped twice, and never re-shipped after a restart.

Query text is normalized before it leaves your host. Literal values are replaced with placeholders, so the stored shape is { userId: "?", status: "?" } rather than your customers' data.

Collection Intervals

Walking every collection every minute is the main way a monitoring agent becomes the problem it was installed to detect — hence the throttles. Both are configurable.

Safety

Watchlog is read-only. It never runs any of the following:

enabling the profiler · dropDatabase / drop · createIndex / dropIndex · compact · reIndex · killOp · setParameter · fsync · any write to any collection

Arguments are passed to mongosh with execFile, never through a shell — so a password containing a quote or ; cannot break out into the command line.

Advanced Options

{
  "service": "mongodb",
  "monitor": true,
  "host": "localhost",
  "port": "27017",
  "username": "watchlog_monitor",
  "password": "your_password",
  "authDatabase": "admin",
  "tls": false,

  "advanced": {
    "enabled": true,
    "storage": true,
    "indexes": true,
    "replication": true,
    "maxDatabases": 50,
    "maxCollectionsPerDatabase": 100,
    "maxCollections": 300,
    "storageIntervalSeconds": 300,
    "indexIntervalSeconds": 300
  },

  "slowQuery": {
    "enabled": false,
    "threshold": 100,
    "maxPerScrape": 100
  }
}

Large deployments

maxDatabases, maxCollectionsPerDatabase and maxCollections bound how much the agent walks per cycle. On a deployment with thousands of collections, raise storageIntervalSeconds before raising the caps — the traversal cost is what matters, not the payload.

Multiple Instances

Add one entry per instance. Each is identified by its host:port:

[
  {
    "service": "mongodb",
    "monitor": true,
    "host": "127.0.0.1",
    "port": "27017",
    "username": "watchlog_monitor",
    "password": "your_password"
  },
  {
    "service": "mongodb",
    "monitor": true,
    "host": "127.0.0.1",
    "port": "27018",
    "username": "watchlog_monitor",
    "password": "your_password"
  }
]

Troubleshooting