← All databases
M

MongoDB monitoring

ClusterEye monitors MongoDB replica sets and sharded clusters through an agent. It connects as a read-only user holding clusterMonitor and readAnyDatabase on the admin database, and collects replica set topology, the oplog window, query performance, and — on sharded clusters — balancer activity and chunk distribution. On a sharded cluster the agent points at mongos. Versions: 4.0 — 7.0.

Agent-basedDefault port: 27017Versions: 4.0 — 7.0
MongoDB — ClusterEye Performance Analyzer

What we read

serverStatus — connections, opcounters, WiredTiger cache, lock and memory metrics
replSetGetStatus and replSetGetConfig — member roles, election history, oplog window and replication lag
currentOp — long-running operations, waiting and blocked work
$collStats and $indexStats — collection size, index usage and unused indexes
system.profile — slow query records (when you have profiling enabled)
the config database and balancerStatus — chunk distribution and balancer state on sharded clusters
dbStats, listDatabases, top — read/write distribution per database and collection
getCmdLineOpts, getParameter and mongod.conf — configuration drift

What it does specifically for MongoDB

Query Intelligence

Slow queries are grouped by fingerprint, prioritised, and given a cost and estimated saving. Patterns doing collection scans (COLLSCAN) and queries that have regressed are flagged separately.

Replica set and oplog visibility

Primary, secondary and arbiter roles on a live map; oplog window, replication lag and election events tracked historically.

Sharded cluster support

Connected through mongos, shards, chunk distribution, balancer activity and jumbo chunks become visible, with unbalanced shards surfaced.

Index usage analysis

$indexStats reveals never-used and overlapping indexes — the ones costing write throughput without helping any read.

Minimum privileges required

The baseline is read-only and safe to grant in production.

javascript
use admin
db.createUser({
  user: "clustereye",
  pwd:  "<strong-password>",
  roles: [
    // serverStatus, replSetGetStatus/Config, listDatabases, dbStats,
    // collStats/$collStats/$indexStats, top, getLog, hostInfo,
    // getCmdLineOpts, getParameter, currentOp, balancerStatus,
    // config.* reads on sharded clusters
    { role: "clusterMonitor",  db: "admin" },

    // system.profile reads, listCollections, index listing, explain
    { role: "readAnyDatabase", db: "admin" }
  ]
})

Prerequisites

·On sharded clusters point the agent at mongos; clusterMonitor already covers config database reads.
·The agent never changes the profiling level, never creates/drops indexes, never runs validate/compact — dbAdminAnyDatabase is not needed.
·OS: read access to the mongod log file and mongod.conf for log analysis and config drift.
Install and privilege details →

Frequently asked questions

Two built-in roles on the admin database: clusterMonitor and readAnyDatabase. dbAdminAnyDatabase is not needed — ClusterEye never changes the profiling level, never creates or drops indexes, and never runs validate or compact.

Start monitoring your MongoDB estate in five minutes.

One command to install. Talk to us if you have questions.

Book a live demoInstall and privilege details