C100DBA Server Administration — 25 Questions

Question 1

Refer to the exhibit. Based on the provided YAML configuration snippet, what will be the behavior of the mongod instance regarding memory allocation on a host with 16GB of total system RAM?

Accepted Answer

The internal cache will be capped strictly at 2GB.. Administrators must understand how to manually override default settings to optimize performance or coexist with other applications. By default, WiredTiger uses 50% of (RAM - 1GB). On a 16GB host, the default would be 7.5GB. However, the configuration explicitly restricts the internal cache to 2GB, which is useful when running multiple instances or memory-constrained containers.

Answer

The instance will use 7.5GB for the WiredTiger cache.

Answer

The total resident memory (RSS) will not exceed 2GB.

Answer

The mongod process will fail to start due to a memory mismatch.

Question 2

A production environment requires point-in-time recovery for a 2TB database. Which TWO strategies are most appropriate for minimizing recovery time while ensuring data consistency?

Accepted Answer

Using filesystem-level snapshots on a volume with journaling enabled.. Point-in-time recovery (PITR) is a fundamental requirement for production database administration. Filesystem snapshots provide a fast way to restore large volumes of data, while the oplog (operations log) allows the DBA to replay changes up to a specific second. Combining these ensures that even a 2TB dataset can be restored quickly without full BSON re-indexing.

Answer

Performing a daily mongodump with the --gzip flag.

Answer

Scheduling a weekly mongorestore to a staging environment.

Answer

Running the repairDatabase command after every write operation.

Question 3

When performing a rolling upgrade of a MongoDB replica set from version 5.0 to 6.0, which TWO steps are mandatory to ensure cluster stability and compatibility?

Accepted Answer

Ensure that 'featureCompatibilityVersion' is set to '5.0' before upgrading.. Upgrading a replica set requires a specific sequence to maintain availability and prevent data corruption. Administrators must ensure that the 'featureCompatibilityVersion' (FCV) matches the old version before starting. Upgrading secondaries first ensures that if a failure occurs, the primary can still function on the old version while the administrator troubleshoots the upgrade process.

Answer

Upgrade the primary node first to ensure new features are available.

Answer

Disable all indexes on the collection before starting the upgrade.

Answer

Run 'db.upgrade()' from the admin database on each node.

Question 4

A database administrator notices that the WiredTiger cache is frequently reaching its maximum size, causing excessive eviction. Which configuration change is the most appropriate to manage memory pressure in a standalone node?

Accepted Answer

Decrease the 'wiredTiger.engineConfig.cacheSizeGB' to a value that allows overhead for the OS.. Tuning the WiredTiger cache size is critical for maintaining performance in memory-constrained environments. By adjusting the 'wiredTiger.engineConfig.cacheSizeGB' parameter, administrators can balance the memory allocated to the engine versus the operating system. This prevents the operating system from swapping to disk, which would otherwise degrade latency. Properly sizing the cache ensures that the working set fits efficiently in RAM, reducing the need for constant page evictions and improving overall throughput.

Answer

Increase the 'wiredTiger.engineConfig.cacheSizeGB' value to consume all available system RAM.

Answer

Enable the 'net.compression.snappy' setting to reduce memory footprint.

Answer

Set 'storage.journal.enabled' to false to free up cache space.

Question 5

An administrator is planning a backup strategy for a replica set. Which TWO of the following methods allow for consistent backups without stopping the MongoDB service?

Accepted Answer

Use 'mongodump' on a secondary node with --oplog enabled.. Maintaining data availability during backups is a primary operational requirement. File system snapshots and MongoDB-native tools like mongodump or dedicated backup solutions are designed for this purpose. Understanding these methods is essential for creating reliable recovery points without causing downtime, which is critical for production environments where continuous availability is mandatory for business continuity and compliance with service level agreements.

Answer

Perform a manual copy of the data files while the server is running.

Answer

Stop the mongod process and copy the contents of the 'dbPath'.

Answer

Use the 'mongorestore' utility to stream data to a remote server.

Question 6

Which THREE of the following actions are standard steps when decommissioning a MongoDB node from a replica set?

Accepted Answer

Execute 'rs.remove()' from the primary node.. Decommissioning a node requires careful coordination to prevent cluster instability. Failing to update the configuration or clean up the node can lead to dangling references or unnecessary connection attempts by other members. Following a strict sequence ensures that the remaining replica set maintains its quorum and that the removed node does not interfere with the health or the security profile of the active production cluster.

Answer

Immediately delete the /var/lib/mongodb directory.

Answer

Force a primary election before removal.

Question 7

What is the effect of setting 'writeConcern' to '{ w: "majority" }' on an application's write operations?

Accepted Answer

The write will persist even if the primary node crashes after acknowledgement.. Setting write concern to 'majority' ensures that the data is acknowledged by a majority of the replica set members before the client receives a success response. This guarantees that the write will not be rolled back during an election, which is critical for maintaining data integrity and preventing 'dirty reads' in distributed systems. It is the best practice for applications requiring strong consistency and durability in a production replica set.

Answer

The write operation will be faster than with 'w: 1'.

Answer

It forces the data to be written directly to the disk cache.

Answer

It automatically enables read-your-writes consistency for all reads.

Question 8

Which MongoDB process utility is specifically used to verify the integrity of the data files and the WiredTiger engine metadata?

Accepted Answer

wiredTiger (standalone utility). Data integrity is paramount in database administration. The 'mongod --repair' or the 'wiredTiger' command-line utility tools allow administrators to validate that data files are not corrupted. In scenarios where a server has experienced a hard crash or power failure, these tools provide the only mechanism to assess the physical health of the data files before attempting to restart the database and risking further data loss or corruption.

Answer

mongodump

Answer

mongofiles

Answer

mongostat

Question 9

An administrator needs to rotate the logs without stopping the MongoDB service. Which approach is the most effective?

Accepted Answer

Use the 'logRotate' command via the mongo shell.. Log rotation is essential for preventing log files from consuming all available disk space. By sending a 'SIGUSR1' signal to the mongod process, administrators can trigger an immediate log rotation. This allows for seamless file management, enabling the archiving of old logs and the creation of new ones without impacting the running database process, which is critical for 24/7 production environments that cannot tolerate downtime.

Answer

Restart the mongod service.

Answer

Manually delete the current log file while the service is running.

Answer

Update the 'systemLog.path' in the config file.

Question 10

A MongoDB replica set has three members: Primary, Secondary1, and Secondary2. Secondary1 is in a different data center with 50 ms latency. You notice that write concern 'majority' is taking longer than expected. Which factor is most likely causing the increased latency for majority writes?

Accepted Answer

Secondary1 must acknowledge the write before the majority write concern is satisfied.. In a three-member replica set, majority write concern requires acknowledgment from two members. If one secondary is unavailable or slow, the primary may need to wait for the other secondary, which has high latency. This directly increases the time to satisfy majority write concern. The other options are incorrect because majority does not require all members or a specific secondary.

Answer

The write concern 'majority' is calculated based on the number of data-bearing members, which excludes arbiters.

Answer

The primary must wait for all secondaries to acknowledge before returning.

Answer

Secondary2 is the only member that can satisfy the majority write concern.

Question 11

You are managing a MongoDB replica set and need to perform maintenance on the primary node. You want to step down the primary gracefully to allow a secondary to take over. Which two commands or methods can you use to achieve this? (Choose two.)

Accepted Answer

Use the replSetStepDown command with a force option.. The two correct methods are rs.stepDown() and the replSetStepDown command with force: true. Both cause the primary to step down gracefully, triggering an election. The other options either shut down the server abruptly, prevent a member from becoming primary, or are invalid commands.

Answer

rs.freeze(120)

Answer

Restart mongod with --shutdown

Answer

db.shutdownServer()

Question 12

A MongoDB DBA needs to restart a mongod process that was started with `--auth` and `--port 27017`, but the original configuration file is unavailable. The DBA must ensure the restarted instance uses the same storage engine (WiredTiger), listens on port 27017, and enforces authentication. Which command will correctly restart the mongod process?

Accepted Answer

mongod --dbpath /data/db --port 27017 --auth --fork. The correct command must restart mongod without a configuration file, explicitly setting the port to 27017 and enabling authentication. The `--fork` flag is also necessary for a typical daemonized restart. The storage engine defaults to WiredTiger in modern MongoDB versions, so no additional flag is needed. This combination ensures the instance runs with the required settings.

Answer

mongod --dbpath /data/db --wiredTigerCacheSizeGB 1 --auth

Answer

mongod --dbpath /data/db --port 27017 --auth

Answer

mongod --config /etc/mongod.conf --fork

Question 13

You are the DBA for a MongoDB 6.0 replica set that runs on Linux servers with a three-member configuration: one primary and two secondaries. The application team reports that a nightly batch job occasionally fails with a 'not primary' error during an automatic failover. You want to reduce the time that clients experience write errors during a primary election without changing the replica set topology. Which server administration action should you take?

Accepted Answer

Lower the electionTimeoutMillis setting in the replica set configuration.. The electionTimeoutMillis parameter determines how long a secondary waits before calling an election when it loses contact with the primary. Lowering it shortens the detection period, so a new primary is elected sooner and clients see fewer 'not primary' errors. The other options either do not affect election timing or change durability and topology without addressing the failover speed.

Answer

Add an arbiter to the replica set to break ties faster.

Answer

Set the writeConcern majority on all application writes.

Answer

Increase the oplog size on all replica set members.

Question 14

A MongoDB 6.0 standalone server is used for development. The server has been running for several months without a restart. A DBA observes that the resident memory reported by the operating system is much higher than the WiredTiger cache size, and the server occasionally becomes unresponsive under load. The DBA wants to reduce memory pressure without changing the hardware. Which administrative action is most appropriate?

Accepted Answer

Restart the mongod process to release accumulated memory.. When mongod's resident memory grows far beyond the WiredTiger cache and the server becomes unresponsive, accumulated memory fragmentation or internal allocations are likely. A restart clears the process memory and restores normal operation. Other options either do not address process memory directly or could worsen the situation. After restart, monitoring and possibly tuning allocator behavior can help prevent recurrence.

Answer

Run the compact command on all collections to release unused space.

Answer

Set the tcmallocReleaseRate parameter to 0 to force immediate memory release.

Answer

Restart mongod with a smaller --wiredTigerCacheSizeGB value.

Question 15

A MongoDB DBA is configuring a new replica set with three members. The DBA wants to ensure that the replica set can tolerate the failure of one member without losing the ability to elect a primary. Which configuration setting must be applied to the members to achieve this?

Accepted Answer

Set `votes` to 1 for all members.. To tolerate the failure of one member in a three-member replica set, all three members must have votes (the default). This ensures that if one member fails, the remaining two still form a majority (2 out of 3) and can elect a primary. The `votes` field controls whether a member can vote in elections; setting it to 1 for all members is essential for this fault tolerance.

Answer

Set `arbiterOnly` to true for one member.

Answer

Set `hidden` to true for one member.

Answer

Set `priority` to 1 for all members.

Question 16

A DBA needs to add a new member to an existing replica set. The new member must be configured as a hidden member with a priority of 0 and a delay of 3600 seconds to serve as a delayed backup. Which sequence of commands should the DBA use?

Accepted Answer

rs.add({ host: 'newhost:27017', hidden: true, priority: 0, slaveDelay: 3600 }). The rs.add command can take a configuration document that includes hidden, priority, and slaveDelay. This single command adds the new member with all desired properties: hidden: true makes it hidden, priority: 0 prevents it from becoming primary, and slaveDelay: 3600 delays replication by one hour. Using rs.add with the full configuration is the correct and efficient method. Other options either use incorrect field names or require a risky multi-step reconfiguration. (Explanation length: 65 words)

Answer

rs.add({ host: 'newhost:27017', hidden: true, priority: 0, secondaryDelaySecs: 3600 })

Answer

rs.add({ host: 'newhost:27017', hidden: true, priority: 0, delay: 3600 })

Answer

rs.add('newhost:27017'); rs.reconfig({ _id: 'rs0', members: [ { _id: 0, host: 'existing1:27017' }, { _id: 1, host: 'existing2:27017' }, { _id: 2, host: 'existing3:27017' }, { _id: 3, host: 'newhost:27017', hidden: true, priority: 0, slaveDelay: 3600 } ] })

Question 17

A DBA is planning to upgrade a sharded cluster from MongoDB 4.4 to 5.0. The cluster has three shards, each a three-member replica set, and two config servers (also a replica set). The DBA must ensure a smooth upgrade with minimal downtime. Which two prerequisites must be met before starting the upgrade? (Choose two.)

Accepted Answer

All shards and config servers must be running MongoDB 4.4.. The two prerequisites for upgrading a sharded cluster from 4.4 to 5.0 are that all shards and config servers must be running MongoDB 4.4, and the featureCompatibilityVersion must be set to "4.4" on all components. These ensure that the cluster is in a consistent state and that the upgrade path is supported. Other steps like stopping the balancer or upgrading clients are recommended but not required.

Answer

Each shard must have at least one secondary with zero replication lag.

Answer

All clients must be upgraded to the MongoDB 5.0 drivers.

Answer

The balancer must be stopped before upgrading any shard.

Question 18

A DBA is configuring a new MongoDB 6.0 replica set for a production environment. The security policy requires that all client connections use TLS/SSL and that the replica set members authenticate each other using x.509 certificates. The DBA has already obtained valid certificates for each server and the clients. Which TWO mongod configuration options are required to enable both TLS for client connections and internal x.509 authentication? (Choose two.)

Accepted Answer

security.clusterAuthMode set to x509. To require TLS for all connections, net.tls.mode must be set to requireTLS. To enable x.509 authentication between replica set members, security.clusterAuthMode must be set to x509. These two settings together satisfy the policy. Other options either support TLS but do not enable it, or relate to client authorization or LDAP, which are not required here.

Answer

security.authorization set to enabled

Answer

net.tls.CAFile set to the CA certificate path

Answer

security.ldap.authz.queryTemplate set to a valid template

Question 19

A MongoDB DBA is investigating a replica set where the primary is experiencing high write latency. The DBA suspects that the write concern is causing delays. Which TWO of the following write concerns would require acknowledgment from a majority of voting members before returning success? (Choose two.)

Accepted Answer

{ w: 2 }. The write concerns that require acknowledgment from a majority of voting members are { w: 'majority' } and, in a typical three-member replica set, { w: 2 } because two out of three is a majority. The w: 'majority' option dynamically adjusts to the number of voting members, while w: 2 is a fixed number that may not always be a majority in larger sets. The other options either do not wait for majority acknowledgment or are invalid. (Explanation length: 70 words)

Answer

{ w: 0 }

Answer

{ w: 1 }

Answer

{ w: 'all' }

Question 20

A DBA needs to automate log rotation on a Linux-based MongoDB server without restarting the mongod process. Which signal or command should be used to ensure the current log file is archived and a new one is started?

Accepted Answer

Send a SIGUSR1 signal to the mongod PID.. Log rotation is a critical administrative task that prevents disk space exhaustion and facilitates log analysis. Using the SIGUSR1 signal on Linux systems allows the administrator to rotate logs without any downtime or server restarts. This ensures continuous availability while maintaining manageable file sizes for diagnostic purposes, which is essential for long-term server health and compliance with retention policies.

Answer

Send a SIGHUP signal to the mongod PID.

Answer

Run the 'rotateLogs' command from the admin database.

Answer

Update the 'systemLog.path' in the YAML config and run 'mongod --reconfig'.