I85PREP
⌘ K
MongoDB DBA Associate study materials
MongoDB DBA Associate study resources

MongoDB DBA Associate practice questions

One question at a time. Build your MongoDB DBA Associate understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Deployment, Installation, and Topology Planning

A team wants a replica set that can survive the loss of any single data center while still electing a primary. What is the minimum number of voting members, and how should they be distributed?

  1. Four voting members split evenly across two data centers
  2. Three voting members spread across three data centers Correct answer.
  3. Five voting members all placed in one data center
  4. Two voting members, one in each of two data centers

Explanation

A replica set needs a strict majority of voting members to elect and maintain a primary, so three members tolerate one failure while leaving two to form a majority. Distributing the three across three data centers ensures a single site outage removes only one vote. Two-member sets have no majority after one loss, and four members split 2-2 cannot elect across a partition.

Question 2Deployment, Installation, and Topology Planning

Which MongoDB deployment topology provides automatic failover with no manual intervention when the primary becomes unavailable?

  1. A replica set with an odd number of voting members Correct answer.
  2. A standalone mongod with journaling enabled
  3. A single mongos router connected to one shard
  4. A sharded cluster with a single config server

Explanation

A replica set elects a new primary automatically when the current primary is lost, provided a majority of voting members remain reachable. A standalone server has no second copy and therefore no failover path. Sharded clusters rely on replica sets internally for that same election behavior.

Question 3Deployment, Installation, and Topology Planning

What is the role of the config server replica set in a sharded cluster?

  1. It holds the oplog used to replicate data between shards
  2. It performs the final merge of results returned by mongos
  3. It stores the chunk distribution and sharding metadata for the cluster Correct answer.
  4. It caches the most frequently queried documents for each shard

Explanation

Config servers persist the cluster's metadata, including which chunks live on which shard and the routing table mongos uses. They must themselves run as a replica set so that metadata remains highly available. They do not store user query results or replicate user data between shards.

Question 4Deployment, Installation, and Topology Planning

Which component must a client application connect to when using a sharded cluster?

  1. A mongos router Correct answer.
  2. The config server primary
  3. A randomly chosen shard's primary
  4. Any secondary in any shard

Explanation

Clients send queries to mongos, which routes them to the appropriate shards and merges the results. Connecting directly to a shard bypasses routing and returns incomplete results for sharded collections. Config servers are internal metadata stores and are not client entry points.

Question 5Deployment, Installation, and Topology Planning

An administrator wants a replica set member that can vote in elections but never holds data or becomes primary. Which member configuration achieves this?

  1. secondaryDelaySecs 3600
  2. arbiterOnly true Correct answer.
  3. priority 0 and votes 1
  4. priority 0 and hidden true

Explanation

An arbiter is a voting member that stores no data and cannot become primary, which lets a set reach an odd voting majority without the cost of a full data-bearing node. Setting priority to 0 alone still stores data and can serve reads. A delayed secondary also stores the full data set.

Question 6Deployment, Installation, and Topology Planning

Which storage engine is the default for MongoDB since version 3.2 and supports document-level concurrency control?

  1. WiredTiger Correct answer.
  2. Encrypted
  3. In-Memory
  4. MMAPv1

Explanation

WiredTiger became the default storage engine in MongoDB 3.2 and provides document-level concurrency plus compression. MMAPv1 was the earlier default and used collection-level locking. The In-Memory engine is an Enterprise option used for specific low-latency workloads.

Question 7Deployment, Installation, and Topology Planning

What does the read preference mode 'primaryPreferred' do for a client connected to a replica set?

  1. It reads only from the primary and fails if the primary is down
  2. It reads from a majority of members and reconciles the results
  3. It always reads from the nearest secondary regardless of primary state
  4. It reads from the primary when available and otherwise from a secondary Correct answer.

Explanation

primaryPreferred routes reads to the primary under normal conditions and falls back to a secondary only when no primary is available. This trades possible staleness during failover for continued read availability. It never reads from multiple members to reconcile results.

Question 8Deployment, Installation, and Topology Planning

Which write concern setting asks MongoDB to acknowledge a write only after it has been applied on a majority of voting members?

  1. w: 1
  2. j: false
  3. w: 0
  4. w: "majority" Correct answer.

Explanation

The write concern w: "majority" waits until a majority of voting members have acknowledged the write, reducing the chance that a failover rolls it back. w: 1 acknowledges only the primary, and w: 0 requests no acknowledgement at all. The journal flag controls durability to disk, not replication count.

Question 9Deployment, Installation, and Topology Planning

A production replica set has three data-bearing members and one arbiter. An administrator wants to add a second arbiter to improve resilience. What is the main risk of this plan?

  1. Adding a second arbiter increases the chance of split votes without adding a data copy Correct answer.
  2. The set would have two arbiters, and MongoDB limits a replica set to one arbiter
  3. Arbiters require the same storage capacity as data-bearing members, doubling cost
  4. Arbiters cannot be added to a replica set that already has data-bearing members

Explanation

Arbiters add votes but no data, so more arbiters can make election outcomes depend on lightweight nodes and increase the risk of split votes and rollbacks without improving durability. MongoDB documentation advises using arbiters sparingly and preferably avoiding them in favor of data-bearing members. Arbiters store no data, so capacity is not the concern.

Question 10Deployment, Installation, and Topology Planning

A sharded cluster uses a hashed shard key on a monotonically increasing field. What is the primary benefit compared with a ranged shard key on the same field?

  1. It keeps range queries targeted to fewer shards
  2. It allows the shard key to be changed without a reshard
  3. It removes the need for a config server replica set
  4. It distributes inserts evenly across shards Correct answer.

Explanation

A hashed shard key scatters values across the key space, so sequential inserts spread over many chunks and shards instead of concentrating on the last chunk. The tradeoff is that range queries typically must be broadcast to all shards. Hashing does not eliminate the config server or the need for resharding to change the key.

Question 11Deployment, Installation, and Topology Planning

Which deployment pattern best supports a workload that must continue serving reads during a primary election in a replica set?

  1. Convert all secondaries to arbiters to speed elections
  2. Disable the oplog to reduce election latency
  3. Set read preference to secondaryPreferred and enable retryable reads Correct answer.
  4. Set read preference to primary and increase the election timeout

Explanation

With secondaryPreferred, reads can continue from secondaries while the election completes, and retryable reads let the driver transparently retry operations interrupted by the election. Reading only from the primary means reads fail during the election window. Arbiters and disabling the oplog do not make reads available during failover.

Question 12Deployment, Installation, and Topology Planning

A team is planning a replica set for a workload with heavy read traffic and wants secondary reads to be consistent with the primary. Which combination is most appropriate?

  1. Read preference secondary with read concern majority and write concern majority Correct answer.
  2. Read preference primary with read concern local
  3. Read preference secondary with write concern w: 1
  4. Read preference nearest with write concern w: 0

Explanation

Reading from secondaries with read concern majority, combined with write concern majority, ensures the data read has been acknowledged by a majority and will not be rolled back. Secondary reads with only w: 1 can return data that a failover later discards. Unacknowledged writes provide no durability guarantee.

Question 13Deployment, Installation, and Topology Planning

What is the purpose of the local database on a replica set member?

  1. It stores the key file used for internal authentication
  2. It stores the cluster's chunk metadata for the sharded cluster
  3. It holds user collections that are excluded from replication for performance
  4. It stores data that is not replicated to other members, such as the oplog and startup diagnostics Correct answer.

Explanation

The local database holds collections that are never replicated, including the oplog, replication diagnostics, and startup logs. Because it is not replicated, it is not included in backups of a replica set taken with mongodump of replicated databases. Chunk metadata lives on config servers, not in local.

Question 14Deployment, Installation, and Topology Planning

An operator needs to run a MongoDB deployment on a single machine for development with the smallest footprint while retaining transactions. Which topology is appropriate?

  1. Two mongod processes without a replica set configuration
  2. A sharded cluster with one shard and one config server
  3. A standalone mongod with no replica set
  4. A single-node replica set Correct answer.

Explanation

Multi-document transactions require a replica set, and a single-node replica set provides that capability with minimal resources on one host. A standalone mongod does not support multi-document transactions. A sharded cluster adds unnecessary mongos and config server overhead for development.

Question 15Deployment, Installation, and Topology Planning

Which authentication mechanism allows a replica set to use a single shared key file for internal member authentication?

  1. LDAP proxy authentication
  2. x.509 certificates
  3. Keyfile authentication Correct answer.
  4. SCRAM-SHA-256

Explanation

Keyfile authentication uses a shared secret file distributed to each member, and members authenticate to each other with it. SCRAM is for client and user authentication, and x.509 uses certificates rather than a shared key file. LDAP proxy auth is an Enterprise feature for external user directories, not internal member auth.

Question 16Deployment, Installation, and Topology Planning

A sharded cluster has three shards, each a three-member replica set. What is the minimum number of voting members required for the config server replica set, and why?

  1. Two, because a majority of two is sufficient for elections
  2. Three, so the metadata store can elect a primary after one failure Correct answer.
  3. Five, because the number must match the total shard count
  4. One, because config data is read-only after startup

Explanation

The config server replica set must be able to elect and maintain a primary, which requires a majority; three members tolerate one failure and still hold a majority of two. A single config server cannot fail over. Two members cannot form a majority after one is lost, and the count need not match the shard count.

Question 17Deployment, Installation, and Topology Planning

An administrator wants backups taken from a secondary to avoid impacting primary performance. Which consideration matters most for consistency?

  1. The secondary must have priority 0 to be eligible for backup
  2. The backup must be taken with w: 0 to avoid locking
  3. The secondary must be hidden for the backup to be valid
  4. The secondary must be caught up and the backup should account for replication lag Correct answer.

Explanation

Backing up a secondary is valid only if the node is reasonably current; replication lag means the backup may omit recent writes. Priority 0 or hidden status affects election and visibility, not backup consistency. Write concern does not apply to a mongodump read operation.

Question 18Deployment, Installation, and Topology Planning

What happens to a replica set with five voting members when two members become unreachable?

  1. The set continues with the existing primary indefinitely
  2. The set automatically converts the unreachable members to arbiters
  3. The set elects a new primary from the three reachable members Correct answer.
  4. The set loses its primary and cannot elect a new one until a majority is reachable

Explanation

Five voting members require a majority of three to elect a primary, and three reachable members satisfy that. The existing primary steps down when it cannot see a majority, and the remaining majority elects a new primary. Unreachable members are not automatically converted to arbiters.

Question 19Deployment, Installation, and Topology Planning

Which statement about connecting to a replica set from an application is correct?

  1. The connection string must list every member including arbiters
  2. The connection string should list the replica set name and multiple seed hosts Correct answer.
  3. The application must connect to the primary hostname only
  4. The application must reconnect manually after every election

Explanation

Including the replicaSet option and multiple seed hosts lets the driver discover the full topology and follow elections automatically. Listing every member is unnecessary, and arbiters are not data-bearing connection targets. Drivers handle reconnection after elections without manual intervention.

Question 20Deployment, Installation, and Topology Planning

A team wants to add a new member to a replica set with minimal impact on the primary. Which approach is recommended?

  1. Add the member with priority 0 and votes 0 permanently
  2. Seed the new member from the primary during peak hours
  3. Stop the primary and copy its data files to the new host
  4. Seed the new member from an existing secondary using initial sync from that member Correct answer.

Explanation

Initial sync pulls a full copy of the data, and sourcing that sync from a secondary keeps the extra read load off the primary. Copying files from a stopped primary causes downtime, and seeding from the primary adds load during peak hours. A member with votes 0 and priority 0 still needs to sync data.

Question 21Deployment, Installation, and Topology Planning

Which factor most directly determines the minimum working set size an administrator should plan for when sizing memory?

  1. The total size of all collections on disk
  2. The size of the oplog collection
  3. The number of mongos routers in the cluster
  4. The indexes and frequently accessed documents that must stay in RAM Correct answer.

Explanation

The working set is the portion of data and indexes actively used by queries, and it should fit in RAM to avoid disk reads. Total on-disk size can greatly exceed the working set and is not the sizing target. Oplog size and router count affect other resources but not the working set.

Question 22Deployment, Installation, and Topology Planning

A replica set must tolerate the loss of two members while still electing a primary. What is the minimum number of voting members?

  1. Three
  2. Seven
  3. Five Correct answer.
  4. Four

Explanation

Tolerating two failures means at least three members must remain to form a majority, so the set needs five voting members (majority of three). With three members, losing two leaves one, which is not a majority. Four members tolerate only one failure because a majority is three.

Question 23Deployment, Installation, and Topology Planning

An operator configures a replica set member with priority 0, hidden true, and votes 1. What behavior results?

  1. It cannot vote in elections and cannot become primary
  2. It cannot be elected primary and is hidden from client reads but still votes Correct answer.
  3. It becomes an arbiter and stores no data
  4. It can be elected primary but is invisible to clients

Explanation

Priority 0 removes the member from primary eligibility, hidden true keeps it out of client read routing, and votes 1 lets it participate in elections. Hidden members still store the full data set and replicate the oplog. Only arbiterOnly makes a member a non-data-bearing non-candidate.

Question 24Deployment, Installation, and Topology Planning

A sharded cluster uses a ranged shard key on a field with low cardinality and skewed distribution. What is the most likely operational consequence?

  1. Loss of the config server replica set majority
  2. Even chunk distribution because the balancer splits chunks by size
  3. Automatic conversion of the shard key to hashed
  4. Jumbo chunks that cannot be split and uneven shard utilization Correct answer.

Explanation

Low-cardinality, skewed shard keys concentrate many documents on few key values, producing chunks that exceed the size limit and cannot be split, known as jumbo chunks. This leaves shards unevenly loaded because the balancer cannot move those chunks. MongoDB does not automatically convert a shard key to hashed.

Question 25Deployment, Installation, and Topology Planning

A replica set has members in two data centers: two members in site A and one member in site B. Site B is lost. What is the resulting state?

  1. The two members in site A elect a primary and continue serving writes Correct answer.
  2. The member in site B is automatically replaced by an arbiter
  3. The remaining members step down and the set halts all operations
  4. The set becomes read-only until site B returns

Explanation

Three voting members require a majority of two, and the two members in site A constitute that majority, so they elect a primary and continue. The risk in this layout is the reverse case: losing site A leaves only one member, which cannot elect. No automatic arbiter replacement occurs.

Question 26Deployment, Installation, and Topology Planning

An administrator installs MongoDB on a Linux server using the official tarball and wants the mongod process to start automatically at boot under a dedicated service account. Which combination of steps accomplishes this on a systemd-based distribution?

  1. Run mongod with the --fork option from a shell profile script so it launches whenever the service account logs in
  2. Start mongod manually and use the db.shutdownServer() command in a cron job to keep it running
  3. Add the mongod binary path to the system PATH and rely on the package manager to register the daemon
  4. Create a systemd unit file that starts mongod with a config file, then enable it with systemctl enable Correct answer.

Explanation

On systemd hosts, the supported way to run mongod as a managed service is to define a unit file that invokes mongod with a configuration file (typically /etc/mongod.conf) and then enable that unit so it starts at boot. The service account and data/log directory ownership must be set correctly for the unit to succeed.

Question 27Deployment, Installation, and Topology Planning

A team is choosing between deploying MongoDB on their own hosts and using a managed Atlas cluster for a new application. Which consideration is a genuine planning difference between these two deployment models?

  1. Self-managed deployments cannot be upgraded across minor MongoDB versions
  2. Only self-managed deployments can use replica sets for high availability
  3. Managed Atlas clusters shift patching and backup scheduling to the provider, while self-managed hosts require the DBA to perform them Correct answer.
  4. Only Atlas clusters can enable authentication for client connections

Explanation

The core operational trade-off is responsibility: with Atlas the provider handles patching, upgrades, and backup automation, while a self-managed deployment leaves those recurring tasks to the DBA team. Both models support replica sets and authentication, so those are not differentiators.

Question 28Deployment, Installation, and Topology Planning

An operator must deploy a replica set whose members authenticate to one another internally, and the security policy forbids storing a shared secret in a plaintext file. Which internal authentication mechanism should be selected?

  1. LDAP proxy authentication against the corporate directory
  2. SCRAM-SHA-256 with per-member user credentials stored in the admin database
  3. x.509 certificates issued by an internal certificate authority Correct answer.
  4. Kerberos with a service principal per member

Explanation

x.509 is the internal authentication mechanism that relies on certificates rather than a shared keyfile, so no plaintext shared secret is stored on disk. Each member presents a certificate signed by a trusted CA, and members validate each other's subject names.

Question 29Deployment, Installation, and Topology Planning

A DBA is planning a two-data-center replica set and wants the deployment to keep a primary available if either single data center is lost. The budget allows exactly five voting members. How should the members be distributed?

  1. Three members in data center A and two members in data center B Correct answer.
  2. Four members in data center A and one member in data center B
  3. Five members split as two in data center A, two in data center B, and one arbiter in a third location
  4. All five members placed in data center A with a delayed secondary in data center B

Explanation

With five voting members split 3/2, losing either site leaves three voters, which is a majority of five and can still elect a primary. A 4/1 split leaves only one voter if site A is lost, so no primary can be elected.

Question 30Deployment, Installation, and Topology Planning

A DBA plans a sharded cluster for a time-series-style collection that grows continuously and is queried mostly by recent time ranges. The team wants chunk migration to stay balanced without hotspots as inserts arrive. Which shard key design best fits these requirements?

  1. A compound shard key of an ascending timestamp plus a high-cardinality tenant identifier Correct answer.
  2. A hashed shard key on the timestamp field
  3. A shard key on a low-cardinality status field
  4. A single-field ranged shard key on the timestamp field

Explanation

A compound key that begins with the monotonically increasing timestamp but includes a high-cardinality field spreads writes across shards while still supporting range queries on time. The mongos balancer can then move chunks without concentrating all new inserts on one shard.

Question 31Replica Set Configuration, Elections, and Failover

In a MongoDB replica set, which member type can accept write operations from clients?

  1. A secondary
  2. A hidden member
  3. An arbiter
  4. A primary Correct answer.

Explanation

Only the primary accepts writes; it records them in the oplog and secondaries replicate them. Secondaries serve reads only when read preference allows, arbiters hold no data, and hidden members are never client-visible.

Question 32Replica Set Configuration, Elections, and Failover

What is the minimum number of voting members required for a MongoDB replica set to elect and maintain a primary?

  1. A strict majority of voting members Correct answer.
  2. Three
  3. One
  4. Two

Explanation

A primary needs a strict majority of voting members to be elected and to stay primary. With three voting members that majority is two, and with five it is three. A single vote is never sufficient because the majority rule scales with set size.

Question 33Replica Set Configuration, Elections, and Failover

Which replica set member stores no data and cannot become primary?

  1. Hidden member
  2. Priority 0 member
  3. Arbiter Correct answer.
  4. Delayed member

Explanation

An arbiter participates in elections with a vote but holds no data set and can never become primary. Priority 0, hidden, and delayed members all store data and simply cannot be elected primary.

Question 34Replica Set Configuration, Elections, and Failover

What happens when a secondary's optime falls so far behind the primary that the primary has already overwritten the oplog entries it needs?

  1. The secondary is removed from the replica set
  2. The primary blocks writes until the secondary catches up
  3. The secondary performs an initial sync Correct answer.
  4. The secondary continues from the oldest available oplog entry

Explanation

When required oplog entries are gone, the secondary cannot catch up incrementally and performs an initial sync, copying the full data set. The primary does not pause writes for a lagging member, and the member is not automatically removed.

Question 35Replica Set Configuration, Elections, and Failover

Which command reports the current replica set configuration, including member priorities and votes?

  1. db.stats()
  2. rs.status()
  3. db.currentOp()
  4. rs.conf() Correct answer.

Explanation

rs.conf() returns the replica set configuration document, including each member's host, priority, votes, and tags. rs.status() reports runtime health rather than the stored configuration.

Question 36Replica Set Configuration, Elections, and Failover

What is the default value of the priority field for a replica set member?

  1. 2
  2. 1 Correct answer.
  3. 0
  4. 10

Explanation

Members default to priority 1, which makes them eligible for election with equal weight. A priority of 0 makes a member permanently ineligible to become primary.

Question 37Replica Set Configuration, Elections, and Failover

What does the write concern option w: "majority" require before a write is acknowledged?

  1. Acknowledgement from the primary only
  2. Acknowledgement from at least one secondary
  3. Acknowledgement from all data-bearing members
  4. Acknowledgement from a majority of voting members Correct answer.

Explanation

w: "majority" waits for acknowledgement from a majority of voting members, which is the same threshold used in elections. This ties durability to the set's ability to elect a new primary.

Question 38Replica Set Configuration, Elections, and Failover

Which read preference directs a client to read only from the current primary?

  1. primaryPreferred
  2. primary Correct answer.
  3. secondary
  4. nearest

Explanation

The primary read preference routes all reads to the current primary and fails if none is available. primaryPreferred falls back to secondaries when no primary is reachable.

Question 39Replica Set Configuration, Elections, and Failover

A three-member replica set has one primary and two secondaries. One secondary fails. What is the resulting state of the set?

  1. The primary steps down and no new primary is elected
  2. The primary remains primary and continues accepting writes Correct answer.
  3. The remaining secondary is promoted to primary
  4. The set becomes read-only until the failed member returns

Explanation

Two of three voting members are still available, which is a majority, so the primary keeps its role and writes continue. A new election is not triggered because the current primary still holds majority support.

Question 40Replica Set Configuration, Elections, and Failover

A three-member replica set loses its primary and one secondary simultaneously. What happens to write availability?

  1. Writes succeed with w:1 but fail with w:majority
  2. Writes continue on the surviving secondary
  3. Writes fail because no majority can elect a primary Correct answer.
  4. Writes continue after a short election delay

Explanation

Only one of three voting members remains, which is not a strict majority, so no primary can be elected and writes fail. Secondaries never accept writes, and write concern cannot change the majority requirement.

Question 41Replica Set Configuration, Elections, and Failover

How can an administrator force the current primary to step down so that another member is elected?

  1. Set the primary's priority to 0 with rs.reconfig()
  2. Restart the primary's mongod process
  3. Run rs.stepDown() on the primary Correct answer.
  4. Run rs.freeze() on the primary

Explanation

rs.stepDown() instructs the primary to relinquish its role, triggering an election among eligible members. rs.freeze() prevents a member from seeking election for a period but does not cause a stepdown.

Question 42Replica Set Configuration, Elections, and Failover

Which condition causes a replica set to hold a new election even though the primary process is still running?

  1. An arbiter is added to the set
  2. A secondary falls behind in oplog application
  3. The primary loses contact with a majority of voting members Correct answer.
  4. A client connects with secondaryPreferred read preference

Explanation

A primary that cannot reach a majority of voting members steps down, and the remaining members elect a new primary. This prevents two primaries from serving writes in a network partition.

Question 43Replica Set Configuration, Elections, and Failover

Why should a replica set have an odd number of voting members?

  1. Odd numbers eliminate the need for arbiters in all cases
  2. Odd numbers avoid tie votes and wasted votes Correct answer.
  3. Odd numbers reduce replication lag
  4. Odd numbers are required by the WiredTiger storage engine

Explanation

With an odd number of voting members, a majority can always be formed without a tie, and each vote carries more weight. An even number can produce split votes and requires an extra member to break ties.

Question 44Replica Set Configuration, Elections, and Failover

What does a member with priority 0 do during an election?

  1. It becomes primary after a fixed timeout
  2. It neither votes nor seeks election
  3. It votes but never seeks election as primary Correct answer.
  4. It seeks election only if all other members are unavailable

Explanation

Priority 0 makes a member ineligible to become primary while it still casts its vote in elections. This is common for members in a remote data center that should serve reads only.

Question 45Replica Set Configuration, Elections, and Failover

How does an arbiter affect the fault tolerance of a three-member replica set with two data-bearing members and one arbiter?

  1. The set can lose both data-bearing members and still accept writes
  2. The set can lose the arbiter and retain all three election votes
  3. The set requires all three members to elect a primary
  4. The set can lose one data-bearing member and still elect a primary Correct answer.

Explanation

The arbiter provides a third vote so that losing one data-bearing member still leaves a majority able to elect a primary. However, the arbiter stores no data, so it does not add a copy of the data set. Majority write concern still requires both data-bearing members, so it cannot be satisfied during that outage.

Question 46Replica Set Configuration, Elections, and Failover

What is the purpose of the heartbeat interval in a replica set?

  1. To determine how often the oplog is compacted
  2. To control how often members exchange status and detect failures Correct answer.
  3. To set the delay before a delayed member applies oplog entries
  4. To set the timeout for client connection pooling

Explanation

Members send heartbeats at a fixed interval so each can track the health of the others. If a primary stops hearing from a majority within the election timeout, it steps down and an election follows.

Question 47Replica Set Configuration, Elections, and Failover

Which replica set configuration field controls how long a member waits before it may seek election after the primary becomes unreachable?

  1. electionTimeoutMillis Correct answer.
  2. slaveDelay
  3. priority
  4. votes

Explanation

electionTimeoutMillis sets the period a member waits without hearing from the primary before calling an election. Lower values speed failover but increase the risk of unnecessary elections during transient network blips.

Question 48Replica Set Configuration, Elections, and Failover

An administrator wants a member to apply oplog entries two hours behind the primary to protect against accidental data deletion. Which configuration accomplishes this?

  1. Set priority to 0
  2. Set slaveDelay to 7200 Correct answer.
  3. Set votes to 0
  4. Set hidden to true

Explanation

slaveDelay sets the number of seconds a member lags behind the primary, so 7200 equals two hours. Such members are typically also hidden and given priority 0 so clients never read stale data from them.

Question 49Replica Set Configuration, Elections, and Failover

What is the effect of setting a member's hidden field to true?

  1. The member stops replicating data
  2. The member can no longer vote in elections
  3. The member is removed from the replica set configuration
  4. The member is excluded from read preference routing and is invisible to clients Correct answer.

Explanation

A hidden member continues to replicate and vote but is not visible to client drivers and is never selected by read preference. It is often combined with priority 0 and a delay for reporting or backup workloads.

Question 50Replica Set Configuration, Elections, and Failover

A replica set has five voting members. What is the minimum number that must be available for a primary to remain in place?

  1. Three Correct answer.
  2. Two
  3. Five
  4. Four

Explanation

A strict majority of five voting members is three, so three available members are enough to elect and maintain a primary. The set can therefore tolerate the loss of two voting members.