MongoDB DBA Associate study plan

Stand up a three-member replica set on separate ports on one host, then verify state transitions by reading rs.status() output member by member

Draw and label the deployment topology options (standalone, replica set, sharded cluster) and write the failure-domain rule for each

Configure an arbiter and a hidden member, then explain in writing when each is appropriate and when it is a liability

Reproduce a primary stepdown with rs.stepDown() and record how long the election takes and which node wins

Configure priorities and member tags, then force elections to observe which member becomes primary under each setting

Simulate a network partition by blocking the replica set port and document how majority and write concern behave

Set readPreference and writeConcern combinations and test which reads return stale data after an intentional failover

Write the exact sequence an operator should follow to safely remove and re-add a member

Deploy a two-shard cluster with config server replica set and verify chunk distribution with sh.status()

Choose shard keys for three sample workloads and justify each against cardinality, monotonicity, and read/write targeting

Trigger the balancer manually and observe chunk migration behavior and its impact on mongos connections

Test a hashed shard key versus a ranged key on the same collection and compare query routing

Compare WiredTiger cache sizing and compression options and set them explicitly in a mongod config file

Use mongostat, mongotop, and db.serverStatus() to identify a slow query and confirm the bottleneck with explain("executionStats")

Model the same one-to-many relationship with embedding and with referencing and measure document size and query cost for both

Create the indexes that support your slowest queries and verify index usage with the winning plan

Enable SCRAM-SHA-256 authentication and create role-based users scoped to specific databases

Configure TLS on all replica set members with a self-signed CA and verify the connection fails without the certificate

Enable encryption at rest with a key file and confirm data files are unreadable without the key

Write the authorization matrix mapping each application role to the minimum required privileges

Take a consistent backup with mongodump against a replica set and restore it into a fresh deployment with mongorestore

Perform a point-in-time recovery using oplog replay and verify the recovered document count matches expectations

Script a recurring backup job with retention and verify the restore path end to end

Compare mongodump, filesystem snapshots, and Ops Manager backup and state when each is the right choice

Configure the profiler at slowms threshold and analyze the slow query log for a full day of synthetic traffic

Run a rolling upgrade of a replica set from one minor version to the next without downtime and document each step

Build a maintenance runbook for index rebuilds and compaction that includes a rollback path

Set up alerting thresholds for replication lag, connections, and disk usage and justify each number

Retake a full-length timed practice set and score each domain separately to find weak areas

Re-drill the two lowest-scoring domains with hands-on labs rather than rereading notes

Recite from memory the commands for replica set initiation, shard key changes, and point-in-time restore

Review every runbook you wrote and fill any gap where a step assumes knowledge you cannot state aloud