SPLK-2002 Splunk Enterprise Certified Architect Questions and Answers

According to the Splunk documentation1, indexer cluster knowledge bundles are the configuration files that the cluster master distributes to the peer nodes as part of the cluster configuration bundle. The knowledge bundles contain the knowledge objects, such as event types, tags, lookups, and so on, that are relevant for indexing and searching the data. The cluster master creates the knowledge bundles by merging the app-name/default and app-name/local directories from the apps that reside on the master node. The cluster master then pushes the knowledge bundles to the peer nodes, where they reside under the $SPLUNK_HOME/var/run directory2. The other options are false because:

Only app-name/local is pushed. This is false because the cluster master pushes both the app-name/default and app-name/local directories, after merging them, to the peer nodes. The app-name/local directory contains the local customizations of the app configuration, while the app-name/default directory contains the default app configuration3.

Only app-name/default is pushed. This is false because the cluster master pushes both the app-name/default and app-name/local directories, after merging them, to the peer nodes. The app-name/default directory contains the default app configuration, while the app-name/local directory contains the local customizations of the app configuration3.

app-name/default and app-name/local are pushed without change. This is false because the cluster master merges the app-name/default and app-name/local directories before pushing them to the peer nodes. This ensures that the peer nodes have the latest and consistent configuration of the apps3.

According to the Splunk blog1, the Universal Forwarder is ideal for collecting data from high-velocity data sources, such as a syslog server, due to its smaller footprint and faster performance. The Universal Forwarder performs minimal processing and sends raw or unparsed data to the indexers, reducing the network traffic and the load on the forwarders. The other options are false because:

When most of the data requires masking, a Heavy Forwarder is needed, as it can perform advanced filtering and data transformation before forwarding the data2.

When data comes directly from a database server, a Heavy Forwarder is needed, as it can run modular inputs such as DB Connect to collect data from various databases2.

When a modular input is needed, a Heavy Forwarder is needed, as the Universal Forwarder does not include a bundled version of Python, which is required for most modular inputs2.

The minimum reference server specification for a Splunk indexer is 12 CPU cores, 12GB RAM, and 800 IOPS. This specification is based on the assumption that the indexer will handle an average indexing volume of 100GB per day, with a peak of 300GB per day, and a typical search load of 1 concurrent search per 1GB of indexing volume. The other specifications are either higher or lower than the minimum requirement. For more information, see [Reference hardware] in the Splunk documentation.

The splunk_instrumentation.log file is the least helpful in troubleshooting a crash, because it contains information about the Splunk Instrumentation feature, which collects and sends usage data to Splunk Inc. for product improvement purposes. This file does not contain any information about the Splunk processes, errors, or crashes. The other options are more helpful in troubleshooting a crash, because they contain relevant information about the Splunk daemon, the standard error output, and the crash report12

1: https://docs.splunk.com/Documentation/Splunk/9.1.2/Troubleshooting/WhatSplunklogsaboutitself#splunk_instrumentation.log 2: https://docs.splunk.com/Documentation/Splunk/9.1.2/Troubleshooting/WhatSplunklogsaboutitself#splunkd_stderr.log

The following statements about integrating with third-party systems are true: You can use Splunk alerts to provision actions on a third-party system, and you can forward data from Splunk forwarder to a third-party system without indexing it first. Splunk alerts are triggered events that can execute custom actions, such as sending an email, running a script, or calling a webhook. Splunk alerts can be used to integrate with third-party systems, such as ticketing systems, notification services, or automation platforms. For example, you can use Splunk alerts to create a ticket in ServiceNow, send a message to Slack, or trigger a workflow in Ansible. Splunk forwarders are Splunk instances that collect and forward data to other Splunk instances, such as indexers or heavy forwarders. Splunk forwarders can also forward data to third-party systems, such as Hadoop, Kafka, or AWS Kinesis, without indexing it first. This can be useful for sending data to other data processing or storage systems, or for integrating with other analytics or monitoring tools. A Hadoop application cannot search data in Splunk, because Splunk does not provide a native interface for Hadoop applications to access Splunk data. Splunk can search data in the Hadoop File System (HDFS), but only by using the Hadoop Connect app, which is a Splunk app that enables Splunk to index and search data stored in HDFS

According to the Splunk documentation1, the Splunk deployment topology is the part of the deployment plan that is vital prior to installing Splunk indexer clusters and search head clusters. The deployment topology defines the number and type of Splunk components, such as forwarders, indexers, search heads, and deployers, that you need to install and configure for your distributed deployment. The deployment topology also determines the network and hardware requirements, the data flow and replication, the high availability and disaster recovery options, and the security and performance considerations for your deployment2. The other options are false because:

Data source inventory is not the part of the deployment plan that is vital prior to installing Splunk indexer clusters and search head clusters, as it is a preliminary step that helps you identify the types, formats, locations, and volumes of data that you want to collect and analyze with Splunk. Data source inventory is important for planning your data ingestion and retention strategies, but it does not directly affect the installation and configuration of Splunk components3.

Data policy definitions are not the part of the deployment plan that is vital prior to installing Splunk indexer clusters and search head clusters, as they are the rules and guidelines that govern how you handle, store, and protect your data. Data policy definitions are important for ensuring data quality, security, and compliance, but they do not directly affect the installation and configuration of Splunk components4.

Education and training plans are not the part of the deployment plan that is vital prior to installing Splunk indexer clusters and search head clusters, as they are the learning resources and programs that help you and your team acquire the skills and knowledge to use Splunk effectively. Education and training plans are important for enhancing your Splunk proficiency and productivity, but they do not directly affect the installation and configuration of Splunk components5.

The _internal index contains license-related events, such as the license usage, the license quota, the license pool, the license stack, and the license violations. These events are logged by the license manager in the license_usage.log file, which is part of the _internal index. The _audit index contains audit events, such as user actions, configuration changes, and search activity. These events are logged by the audit trail in the audit.log file, which is part of the _audit index. The _license index does not exist in Splunk, as the license-related events are stored in the _internal index. The _introspection index contains platform instrumentation data, such as the resource usage, the disk objects, the search activity, and the data ingestion. These data are logged by the introspection generator in various log files, such as resource_usage.log, disk_objects.log, search_activity.log, and data_ingestion.log, which are part of the _introspection index. For more information, see About Splunk Enterprise logging and [About the _internal index] in the Splunk documentation.

According to the Splunk documentation1, multi-site clustering is an indexer cluster that spans multiple physical sites, such as data centers. Each site has its own set of peer nodes and search heads. Each site also obeys site-specific replication and search factor rules. The use cases that are made possible by multi-site clustering are:

Greatly reduce WAN traffic by preferentially searching assigned site (search affinity). This means that if you configure each site so that it has both a search head and a full set of searchable data, the search head on each site will limit its searches to local peer nodes. This eliminates any need, under normal conditions, for search heads to access data on other sites, greatly reducing network traffic between sites2.

Seamlessly route searches to a redundant site in case of a site failure. This means that by storing copies of your data at multiple locations, you maintain access to the data if a disaster strikes at one location. Multisite clusters provide site failover capability. If a site goes down, indexing and searching can continue on the remaining sites, without interruption or loss of data2.

The other options are false because:

Use blockchain technology to audit search activity from geographically dispersed data centers. This is not a use case of multi-site clustering, as Splunk does not use blockchain technology to audit search activity. Splunk uses its own internal logs and metrics to monitor and audit search activity3.

Enable a forwarder to send data to multiple indexers. This is not a use case of multi-site clustering, as forwarders can send data to multiple indexers regardless of whether they are in a single-site or multi-site cluster. This is a basic feature of forwarders that allows load balancing and high availability of data ingestion4.

Of the following types of files within an index bucket, the rawdata file type may consume the most disk. The rawdata file type contains the compressed and encrypted raw data that Splunk has ingested. The rawdata file type is usually the largest file type in a bucket, because it stores the original data without any filtering or extraction. The bloom filter file type contains a probabilistic data structure that is used to determine if a bucket contains events that match a given search. The bloom filter file type is usually very small, because it only stores a bit array of hashes. The metadata (.data) file type contains information about the bucket properties, such as the earliest and latest event timestamps, the number of events, and the size of the bucket. The metadata file type is also usually very small, because it only stores a few lines of text. The inverted index (.tsidx) file type contains the time-series index that maps the timestamps and event IDs of the raw data. The inverted index file type can vary in size depending on the number and frequency of events, but it is usually smaller than the rawdata file type

The Splunk Enterprise Security and Encryption documentation specifies that the primary mechanism for securing data in motion within a Splunk environment is to enable TLS/SSL encryption between forwarders and indexers. This ensures that log data transmitted from Universal Forwarders or Heavy Forwarders to Indexers is fully encrypted and protected from interception or tampering.

The correct configuration involves setting up signed SSL certificates on both forwarders and indexers:

On the forwarder, TLS settings are defined in outputs.conf, specifying parameters like sslCertPath, sslPassword, and sslRootCAPath.

On the indexer, TLS is enabled in inputs.conf and server.conf using the same shared CA for validation.

Splunk’s documentation explicitly states that this configuration protects data-in-transit between the collection (forwarder) and indexing (storage) tiers — which is the critical link where sensitive log data is most vulnerable.

Other communication channels (e.g., deployment server to clients or browser to Splunk Web) can also use encryption but do not secure the ingestion pipeline that handles the indexed data stream. Therefore, TLS should be implemented between Splunk forwarders and indexers.

References (Splunk Enterprise Documentation):

• Securing Data in Transit with SSL/TLS

• Configure Forwarder-to-Indexer Encryption Using SSL Certificates

• Server and Forwarder Authentication Setup Guide

• Splunk Enterprise Admin Manual – Security and Encryption Best Practices

According to the Splunk Enterprise Capacity Planning and Hardware Sizing Guidelines, a distributed Splunk environment’s minimum search head specification must ensure that the system can efficiently manage search parsing, ad-hoc query execution, and knowledge object replication. Splunk officially recommends using a 64-bit x86 architecture system with a minimum of two physical CPU cores (or four vCPUs) running at 2 GHz or higher per core for acceptable performance.

Search heads are CPU-intensive components, primarily constrained by processor speed and the number of concurrent searches they must handle. Memory and disk space should scale with user concurrency and search load, but CPU capability remains the baseline requirement. While 128 GB RAM (Option C) is suitable for high-demand or Enterprise Security (ES) deployments, it exceeds the minimum hardware specification for general distributed search environments.

Splunk no longer supports 32-bit architectures (Option B). While a 1Gb Ethernet NIC (Option A) is common, it is not part of the minimum computational specification required by Splunk for search heads. The critical specification is processor capability — two physical cores or equivalent.

References (Splunk Enterprise Documentation):

• Splunk Enterprise Capacity Planning Manual – Hardware and Performance Guidelines

• Search Head Sizing and System Requirements

• Distributed Deployment Manual – Recommended System Specifications

• Splunk Hardware and Performance Tuning Guide

The best practice for ingesting syslog data from network devices on port 514 into Splunk is to configure syslog to write logs and use a Splunk forwarder to collect the logs. This practice will ensure that the data is reliably collected and forwarded to Splunk, without losing any data or overloading the Splunk indexer. Configuring syslog to send the data to multiple Splunk indexers will not guarantee data reliability, as syslog is a UDP protocol that does not provide acknowledgment or delivery confirmation. Using a Splunk indexer to collect a network input on port 514 directly will not provide data reliability or load balancing, as the indexer may not be able to handle the incoming data volume or distribute it to other indexers. Using a Splunk forwarder to collect the input on port 514 and forward the data will not provide data reliability, as the forwarder may not be able to receive the data from syslog or buffer it in case of network issues. For more information, see [Get data from TCP and UDP ports] and [Best practices for syslog data] in the Splunk documentation.

When adding or decommissioning a member from a Search Head Cluster (SHC), the proper order of operations is:

Delete Splunk Enterprise, if it exists.

Install and initialize the instance.

Join the SHC.

This order of operations ensures that the member has a clean and consistent Splunk installation before joining the SHC. Deleting Splunk Enterprise removes any existing configurations and data from the instance. Installing and initializing the instance sets up the Splunk software and the required roles and settings for the SHC. Joining the SHC adds the instance to the cluster and synchronizes the configurations and apps with the other members. The other order of operations are not correct, because they either skip a step or perform the steps in the wrong order.

According to the Splunk documentation1, a bucket within a clustered index contains two key types of files: the raw data in compressed form (rawdata) and the indexes that point to the raw data (tsidx files). A bucket can be either replicated or searchable, depending on whether it has both types of files or only the rawdata file. A replicated bucket is a bucket that has been copied from one peer node to another for the purpose of data replication. A searchable bucket is a bucket that has both the rawdata and the tsidx files, and can be searched by the search heads. The types of files that exist in a bucket within a clustered index are:

Inside a searchable bucket, there is tsidx and rawdata. This is true because a searchable bucket contains both the data and the index files, and can be searched by the search heads1.

Inside a replicated bucket, there is both tsidx and rawdata. This is true because a replicated bucket can also be a searchable bucket, if it has both the data and the index files. However, not all replicated buckets are searchable, as some of them might only have the rawdata file, depending on the replication factor and the search factor settings1.

The other options are false because:

Inside a replicated bucket, there is only rawdata. This is false because a replicated bucket can also have the tsidx file, if it is a searchable bucket. A replicated bucket only has the rawdata file if it is a non-searchable bucket, which means that it cannot be searched by the search heads until it gets the tsidx file from another peer node1.

Inside a searchable bucket, there is only tsidx. This is false because a searchable bucket always has both the tsidx and the rawdata files, as they are both required for searching the data. A searchable bucket cannot exist without the rawdata file, as it contains the actual data that the tsidx file points to1.

As a best practice, the internal licensing logs should be stored on the license server. The license server is a Splunk instance that manages the distribution and enforcement of licenses in a Splunk deployment. The license server generates internal licensing logs that contain information about the license usage, violations, warnings, and pools. The internal licensing logs should be stored on the license server itself, because they are relevant to the license server’s role and function. Storing the internal licensing logs on the license server also simplifies the license monitoring and troubleshooting process. The internal licensing logs should not be stored on the indexing layer, the deployment layer, or the search head layer, because they are not related to the roles and functions of these layers. Storing the internal licensing logs on these layers would also increase the network traffic and disk space consumption

According to the Splunk documentation1, when migrating an indexer cluster from single-site to multi-site, you must remove the existing single-site attributes from the server.conf file of each peer node. These attributes include replication_factor, search_factor, and cluster_label. You must also restart each peer node after removing the attributes. The other options are false because:

Multi-site policies will apply only to the data created after migration, unless you configure the manager node to convert legacy buckets to multi-site1.

All peer nodes do not need to run the same version of Splunk, as long as they are compatible with the manager node2.

Single-site buckets can be converted to multi-site buckets by changing the constrain_singlesite_buckets setting in the manager node’s server.conf file to " false " 1.

The client filters available in serverclass.conf are DNS name, IP address, and platform (machine type). These filters allow the administrator to specify which forwarders belong to a server class and receive the apps and configurations from the deployment server. The Splunk server role is not a valid client filter in serverclass.conf, as it is not a property of the forwarder. For more information, see [Use forwarder management filters] in the Splunk documentation.

inputs.conf is a configuration file that contains settings for various types of data inputs, such as files, directories, network ports, scripts, and so on1.

initCrcLength is a setting that specifies the number of characters that the input uses to calculate the CRC (cyclic redundancy check) of a file1. The CRC is a value that uniquely identifies a file based on its content2.

crcSalt is another setting that adds a string to the CRC calculation to force the input to consume files that have matching CRCs1. This can be useful when files have identical headers or when files are renamed or rolled over2.

When troubleshooting a situation where some files within a directory are not being indexed, the ignored files are discovered to have long headers, the first thing that should be added to inputs.conf is to increase the value of initCrcLength. This is because by default, the input only performs CRC checks against the first 256 bytes of a file, which means that files with long headers may have matching CRCs and be skipped by the input2. By increasing the value of initCrcLength, the input can use more characters from the file to calculate the CRC, which can reduce the chances of CRC collisions and ensure that different files are indexed3.

Option C is the correct answer because it reflects the best practice for troubleshooting this situation. Option A is incorrect because decreasing the value of initCrcLength would make the CRC calculation less reliable and more prone to collisions. Option B is incorrect because adding a crcSalt with a static string would not help differentiate files with long headers, as they would still have matching CRCs. Option D is incorrect because adding a crcSalt with the < SOURCE > attribute would add the full directory path to the CRC calculation, which would not help if the files are in the same directory2.

The master node distributes configuration bundles to peer nodes in the slave-apps directory under $SPLUNK_HOME/etc. The configuration bundle method is the only supported method for managing common configurations and app deployment across the set of peers. It ensures that all peers use the same versions of these files1. Bundles typically contain a subset of files (configuration files and assets) from $SPLUNK_HOME/etc/system, $SPLUNK_HOME/etc/apps, and $SPLUNK_HOME/etc/users2. The process of distributing knowledge bundles means that peers by default receive nearly the entire contents of the search head’s apps3.

In the context of splunkd.log events written to the _internal index, the field that identifies the specific log channel is the " channel " field. This information is confirmed by the Splunk Common Information Model (CIM) documentation, where " channel " is listed as a field name associated with Splunk Audit Logs​​.

 Configurations from the deployer are merged into the SPLUNK_HOME/etc/apps/APP_HOME/local directory on the search head cluster member. The deployer distributes apps and other configurations to the search head cluster members in the form of a configuration bundle. The configuration bundle contains the contents of the SPLUNK_HOME/etc/shcluster/apps directory on the deployer. When a search head cluster member receives the configuration bundle, it merges the contents of the bundle into its own SPLUNK_HOME/etc/apps directory. The configurations in the local directory take precedence over the configurations in the default directory. The SPLUNK_HOME/etc/system/local directory is used for system-level configurations, not app-level configurations. The SPLUNK_HOME/etc/apps/search/default directory is used for the default configurations of the search app, not the configurations from the deployer.

 The parallelIngestionPipelines setting can be adjusted on the indexers and forwarders to improve Splunk performance. The parallelIngestionPipelines setting determines how many concurrent data pipelines are used to process the incoming data. Increasing the parallelIngestionPipelines setting can improve the data ingestion and indexing throughput, especially for high-volume data sources. The parallelIngestionPipelines setting can be adjusted on the indexers and forwarders by editing the limits.conf file. The parallelIngestionPipelines setting cannot be adjusted on the search head or the cluster master, because they are not involved in the data ingestion and indexing process.

This is a mandatory Splunk component when implementing a search head cluster, as it is responsible for distributing the configuration updates and app bundles to the cluster members1. The deployer is a separate instance that communicates with the cluster manager and pushes the changes to the search heads1. The other options are not mandatory components for a search head cluster. Option A, Captain Server, is not a component, but a role that is dynamically assigned to one of the search heads in the cluster2. The captain coordinates the replication and search activities among the cluster members2. Option C, Cluster Manager, is a component for an indexer cluster, not a search head cluster3. The cluster manager manages the replication and search factors, and provides a web interface for monitoring and managing the indexer cluster3. Option D, RAFT Server, is not a component, but a protocol that is used by the search head cluster to elect the captain and maintain the cluster state4. Therefore, option B is the correct answer, and options A, C, and D are incorrect.

1: Use the deployer to distribute apps and configuration updates 2: About the captain 3: About the cluster manager 4: How a search head cluster works

The splunk offline --enforce-counts command is the official and documented method used to gracefully decommission a search peer (indexer) from an indexer cluster in Splunk Enterprise. This command ensures that all replication and search factors are maintained before the peer is removed.

When executed, Splunk initiates a controlled shutdown process for the peer node. The Cluster Manager verifies that sufficient replicated copies of all bucket data exist across the remaining peers according to the configured replication_factor (RF) and search_factor (SF). The --enforce-counts flag specifically enforces that replication and search counts remain intact before the peer fully detaches from the cluster, ensuring no data loss or availability gap.

The sequence typically includes:

Validating cluster state and replication health.

Rolling off the peer’s data responsibilities to other peers.

Removing the peer from the active cluster membership list once replication is complete.

Other options like disablepeer, decommission, or remove cluster-peers are not valid Splunk commands. Therefore, the correct documented method is to use:

splunk offline --enforce-counts

References (Splunk Enterprise Documentation):

• Indexer Clustering: Decommissioning a Peer Node

• Managing Peer Nodes and Maintaining Data Availability

• Splunk CLI Command Reference – splunk offline

• Cluster Manager and Peer Maintenance Procedures

According to the Splunk Search Head Clustering Troubleshooting Guide, when a Search Head Cluster (SHC) member is reverted from a backup or experiences configuration drift (e.g., an outdated Raft state), it can fail to rejoin the cluster due to inconsistent Raft metadata. The Raft database stores the SHC’s internal consensus and replication state, including knowledge object synchronization, captain election history, and peer membership information.

If this Raft metadata becomes corrupted or outdated (as in the scenario where a node is restored from backup), the recommended and Splunk-supported remediation is to clean the Raft metadata using:

splunk clean raft

This command resets the node’s local Raft state so it can re-synchronize with the current SHC captain and rejoin the cluster cleanly.

The steps generally are:

Stop the affected SHC member.

Run splunk clean raft on that node.

Restart Splunk.

Verify that it successfully rejoins the SHC.

Deleting configuration stanzas or forcing re-addition (Options B and C) can lead to further inconsistency or data loss. Reviewing logs (Option A) helps diagnose issues but does not resolve Raft corruption.

References (Splunk Enterprise Documentation):

• Troubleshooting Raft Metadata Corruption in Search Head Clusters

• splunk clean raft Command Reference

• Search Head Clustering: Recovering from Backup and Membership Failures

• Splunk Enterprise Admin Manual – Raft Consensus and SHC Maintenance

Number of concurrent users: This is an important factor because it affects the search performance and resource utilization of the Splunk environment. More users mean more concurrent searches, which require more CPU, memory, and disk I/O. The number of concurrent users also determines the search head capacity and the search head clustering configuration12

Volume of incoming data: This is another crucial factor because it affects the indexing performance and storage requirements of the Splunk environment. More data means more indexing throughput, which requires more CPU, memory, and disk I/O. The volume of incoming data also determines the indexer capacity and the indexer clustering configuration13

Existence of premium apps: This is a relevant factor because some premium apps, such as Splunk Enterprise Security and Splunk IT Service Intelligence, have additional requirements and recommendations for the Splunk environment. For example, Splunk Enterprise Security requires a dedicated search head cluster and a minimum of 12 CPU cores per search head. Splunk IT Service Intelligence requires a minimum of 16 CPU cores and 64 GB of RAM per search head45

The Splunk deployer sends apps to the search head cluster members by default to the path etc/shcluster/ < app-name > /default. The deployer is a Splunk component that distributes apps and configurations to members of a search head cluster.

Splunk ' s documentation recommends placing the configuration bundle in the $SPLUNK_HOME/etc/shcluster/apps directory on the deployer, which then gets distributed to the search head cluster members. However, it should be noted that within each app ' s directory, configurations can be under default or local subdirectories, with local taking precedence over default for configurations. The reference to etc/shcluster/ < app-name > /default is not a standard directory structure and might be a misunderstanding. The correct path where the deployer pushes configuration bundles is $SPLUNK_HOME/etc/shcluster/apps

A single-site indexer cluster is a group of Splunk Enterprise instances that index and replicate data across the cluster1. A bucket is a directory that contains indexed data, along with metadata and other information2. A replication factor is the number of copies of each bucket that the cluster maintains1. A search factor is the number of searchable copies of each bucket that the cluster maintains1. A searchable copy is a copy that contains both the raw data and the index files3. A search head is a Splunk Enterprise instance that coordinates the search activities across the peer nodes1.

Option D is the correct answer because it reflects the definitions of replication factor and search factor. The cluster will ensure that there are at least three copies of each bucket, one on each peer node, to satisfy the replication factor of 3. The cluster will also ensure that there are at least two searchable copies of each bucket, one primary and one searchable, to satisfy the search factor of 2. The primary copy is the one that the search head uses to run searches, and the searchable copy is the one that can be promoted to primary if the original primary copy becomes unavailable3.

Option A is incorrect because it confuses the replication factor and the search factor. The cluster will ensure there are at least three copies of each bucket, not two, to meet the replication factor of 3. The cluster will ensure there are at least two copies of searchable metadata, not three, to meet the search factor of 2.

Option B is incorrect because it uses the wrong terms. The cluster will ensure there are at least, not at most, three copies of each bucket, to meet the replication factor of 3. The cluster will ensure there are at least, not at most, two copies of searchable metadata, to meet the search factor of 2.

Option C is incorrect because it has nothing to do with the replication factor or the search factor. The cluster does not limit the number of search heads that can access the bucket at the same time. The search head can search across multiple clusters, and the cluster can serve multiple search heads1.

1: The basics of indexer cluster architecture - Splunk Documentation 2: About buckets - Splunk Documentation 3: Search factor - Splunk Documentation

According to the Indexer Clustering Administration Guide, configuration bundles pushed from the Cluster Manager (Master Node) overwrite the contents of the $SPLUNK_HOME/etc/slave-apps/ directory on each search peer (indexer). However, Splunk provides a special persistent location — the _cluster app’s local directory — for files that must survive bundle redeployments.

Specifically, any configuration files placed in:

$SPLUNK_HOME/etc/slave-apps/_cluster/local/

will persist after future bundle pushes because this directory is excluded from the automatic overwrite process.

This is particularly useful for maintaining local overrides or custom configurations that should not be replaced by the Cluster Manager, such as environment-specific inputs, temporary test settings, or monitoring configurations unique to that peer.

Other directories under slave-apps are overwritten each time a configuration bundle is pushed, ensuring consistency across the cluster. Likewise, master-apps exists only on the Cluster Manager and is used for deployment, not persistence.

Thus, the _cluster/local folder is the only safe, Splunk-documented location for configurations that need to survive bundle redeployment.

References (Splunk Enterprise Documentation):

• Indexer Clustering: How Configuration Bundles Work

• Maintaining Local Configurations on Clustered Indexers

• slave-apps and _cluster App Structure and Behavior

• Splunk Enterprise Admin Manual – Cluster Configuration Management Best Practices

 The following artifacts are included in a Splunk diag file:

Internal logs. These are the log files that Splunk generates to record its own activities, such as splunkd.log, metrics.log, audit.log, and others. These logs can help troubleshoot Splunk issues and monitor Splunk performance.

Configuration files. These are the files that Splunk uses to configure various aspects of its operation, such as server.conf, indexes.conf, props.conf, transforms.conf, and others. These files can help understand Splunk settings and behavior. The following artifacts are not included in a Splunk diag file:

OS settings. These are the settings of the operating system that Splunk runs on, such as the kernel version, the memory size, the disk space, and others. These settings are not part of the Splunk diag file, but they can be collected separately using the diag --os option.

Customer data. These are the data that Splunk indexes and makes searchable, such as the rawdata and the tsidx files. These data are not part of the Splunk diag file, as they may contain sensitive or confidential information. For more information, see Generate a diagnostic snapshot of your Splunk Enterprise deployment in the Splunk documentation.

In the deployment planning process, a person should identify who gets to see network data in the data policy definition step. This step involves defining the data access policies and permissions for different users and roles in Splunk. The deployment schedule step involves defining the timeline and milestones for the deployment project. The topology diagramming step involves creating a visual representation of the Splunk architecture and components. The data source inventory step involves identifying and documenting the data sources and types that will be ingested by Splunk

The splunk offline --enforce-counts command will permanently decommission a peer node operating in an indexer cluster. This command will remove the peer node from the cluster and delete its data. This command should be used when the peer node is no longer needed or is being replaced by another node. The splunk stop -f command will stop the Splunk service on the peer node, but it will not decommission it from the cluster. The splunk offline -f command will take the peer node offline, but it will not delete its data or enforce the replication and search factors. The splunk decommission --enforce-counts command is not a valid Splunk command. For more information, see Remove a peer node from an indexer cluster in the Splunk documentation.

The __introspection log file contains data about the impact of the Splunk software on the host system, such as CPU, memory, disk, and network usage, as well as KV store performance1. This log file is monitored by default and the contents are sent to the _introspection index1. The other options are not related to the __introspection log file. File monitor input configurations are stored in inputs.conf2. File monitor checkpoint offset is stored in fishbucket3. User activities and knowledge objects are stored in the _audit and _internal indexes respectively4.

A dedicated deployment server is a Splunk instance that manages the distribution of configuration updates and apps to a set of deployment clients, such as forwarders, indexers, or search heads. A dedicated deployment server should be used when there are more than 50 deployment clients, because this number exceeds the recommended limit for a non-dedicated deployment server. A non-dedicated deployment server is a Splunk instance that also performs other roles, such as indexing or searching. Using a dedicated deployment server can improve the performance, scalability, and reliability of the deployment process. Option C is the correct answer. Option A is incorrect because the number of search peers does not affect the need for a dedicated deployment server. Search peers are indexers that participate in a distributed search. Option B is incorrect because the number of apps to deploy does not affect the need for a dedicated deployment server. Apps are packages of configurations and assets that provide specific functionality or views in Splunk. Option D is incorrect because the number of server classes does not affect the need for a dedicated deployment server. Server classes are logical groups of deployment clients that share the same configuration updates and apps12

1: https://docs.splunk.com/Documentation/Splunk/9.1.2/Updating/Aboutdeploymentserver 2: https://docs.splunk.com/Documentation/Splunk/9.1.2/Updating/Whentousedeploymentserver

The correct Splunk deployment to have the recommended minimum components for a high-availability search head cluster is 3 search heads, 1 deployer, 3 indexers. This configuration ensures that the search head cluster has at least three members, which is the minimum number required for a quorum and failover1. The deployer is a separate instance that manages the configuration updates for the search head cluster2. The indexers are the nodes that store and index the data, and having at least three of them provides redundancy and load balancing3. The other options are not recommended, as they either have less than three search heads or less than three indexers, which reduces the availability and reliability of the cluster. Therefore, option B is the correct answer, and options A, C, and D are incorrect.

1: About search head clusters 2: Use the deployer to distribute apps and configuration updates 3: About indexer clusters and index replication

Splunk Enterprise officially requires a minimum of three search heads and one deployer for a supported Search Head Cluster (SHC) configuration. This ensures both high availability and data consistency within the cluster.

The Splunk documentation explains that a search head cluster uses RAFT-based consensus to elect a captain responsible for managing configuration replication, scheduling, and user workload distribution. The RAFT protocol requires a quorum of members to maintain consistency. In practical terms, this means a minimum of three members (search heads) to achieve fault tolerance — allowing one member to fail while maintaining operational stability.

The deployer is a separate Splunk instance responsible for distributing configuration bundles (apps, settings, and user configurations) to all members of the search head cluster. The deployer is not part of the SHC itself but is mandatory for its proper management.

Running with fewer than three search heads or replacing the deployer with a Deployment Server (as in Options B, C, or D) is unsupported and violates Splunk best practices for SHC resiliency and management.

References (Splunk Enterprise Documentation):

• Search Head Clustering Overview – Minimum Supported Architecture

• Deploy and Configure the Deployer for a Search Head Cluster

• High Availability and Fault Tolerance with RAFT in SHC

The splunk edit licenser-localslave command is used to convert a Splunk instance to a license slave. This command will configure the Splunk instance to contact a license master and receive a license from it. This command should be used when the Splunk instance is part of a distributed deployment and needs to share a license pool with other instances. The splunk add licenses command is used to add a license to a Splunk instance, not to convert it to a license slave. The splunk list licenser-slaves command is used to list the license slaves that are connected to a license master, not to convert a Splunk instance to a license slave. The splunk list licenser-localslave command is used to list the license master that a license slave is connected to, not to convert a Splunk instance to a license slave. For more information, see Configure license slaves in the Splunk documentation.

The walklex command in Splunk is a specialized administrative search command used to translate and display LISPY (Splunk’s internal representation of search terms). LISPY is the logical search syntax Splunk uses to parse and execute search queries, and examining it helps administrators and developers debug search optimization, field extraction behavior, and index-time search efficiency.

When you run the command | walklex search= " your_search_string " , Splunk outputs how it tokenizes and interprets that query internally. This is particularly useful for understanding how Splunk’s search language maps to index-time fields and for diagnosing performance issues caused by inefficient search term parsing.

For example:

| walklex search= " error OR failure host=server01 "

Displays the corresponding LISPY translation used by Splunk’s search subsystem.

Other options are unrelated:

dbinspect provides index bucket metadata.

Monitoring Console shows performance metrics and health status.

Search Job Inspector analyzes search execution phases but doesn’t expose LISPY.

Thus, the correct and Splunk-documented tool for LISPY inspection is the walklex command.

References (Splunk Enterprise Documentation):

• walklex Command Reference – LISPY and Search Debugging

• Understanding Search Language Parsing in Splunk

• Search Internals: How Splunk Interprets Queries

• Splunk Search Performance Troubleshooting Tools

The KV store forms its own cluster within a SHC. The maximum number of SHC members KV store will form is 50. The KV store cluster is a subset of the SHC members that are responsible for replicating and storing the KV store data. The KV store cluster can have up to 50 members, but only 20 of them can be active at any given time. The other members are standby members that can take over if an active member fails. The KV store cluster cannot have more than 50 members, nor can it have an unlimited number of members. The KV store cluster cannot have 25 or 100 members, because these numbers are not multiples of 5, which is the minimum replication factor for the KV store cluster

Migration from single-site to multisite index replication only affects new data, not existing data. Multisite policies apply to new data only, meaning that data that is ingested after the migration will follow the multisite replication and search factors. Existing data, or data that was ingested before the migration, will retain the single-site policies, unless they are manually converted to multisite buckets. Single-site buckets do not instantly receive the multisite policies, nor do they automatically convert to multisite buckets. Multisite total values can exceed any single-site factors, as long as they do not exceed the number of peer nodes in the cluster. A master node is not required at each site, only one master node is needed for the entire cluster

 Changing the limits.conf value for max_searches_per_cpu to a higher value is the best option to increase scheduled search capacity on the search head cluster when a large number of searches are skipped across time. This value determines how many concurrent scheduled searches can run on each CPU core of the search head. Increasing this value will allow more scheduled searches to run at the same time, which will reduce the number of skipped searches. Creating a job server on the cluster, running the server.conf captain_is_adhoc_searchhead = true command, or adding another search head to the cluster are not the best options to increase scheduled search capacity on the search head cluster. For more information, see [Configure limits.conf] in the Splunk documentation.

According to the Splunk Search Job Inspector documentation, the search.log file for a completed search job can be accessed through Splunk Web by navigating to the job’s detailed inspection view.

To access it:

Open the completed search in Splunk Web.

Click the Job menu (top right of the search interface).

Select Inspect Job.

In the Job Inspector window, click the search.log link.

This log provides in-depth diagnostic details about how the search was parsed, distributed, and executed across the search head and indexers. It contains valuable performance metrics, command execution order, event sampling information, and any error or warning messages encountered during search processing.

The search.log file is generated for every search job—scheduled, ad-hoc, or background—and is stored in the job’s dispatch directory ($SPLUNK_HOME/var/run/splunk/dispatch/ < search_id > /).

Other listed options are incorrect:

Option A queries the _internal index, which does not store per-search logs.

Option B is used to view search configurations, not logs.

Option C is not a valid Splunk Web navigation option.

Thus, the only correct and Splunk-documented method is via Job → Inspect Job → search.log.

References (Splunk Enterprise Documentation):

• Search Job Inspector Overview and Usage

• Analyzing Search Performance Using search.log

• Search Job Management and Dispatch Directory Structure

• Splunk Enterprise Admin Manual – Troubleshooting Searches

 The Monitoring Console is the Splunk tool that offers a health check for administrators to evaluate the health of their Splunk deployment. The Monitoring Console provides dashboards and alerts that show the status and performance of various Splunk components, such as indexers, search heads, forwarders, license usage, and search activity. The Monitoring Console can also run health checks on the deployment and identify any issues or recommendations. The btool is a command-line tool that shows the effective settings of the configuration files, but it does not offer a health check. The DiagGen is a tool that generates diagnostic snapshots of the Splunk environment, but it does not offer a health check. The SPL Clinic is a tool that analyzes and optimizes SPL queries, but it does not offer a health check. For more information, see About the Monitoring Console in the Splunk documentation.

Index files (. tsidx files) are the main components of an index that store the raw data and the inverted index of terms. They take the most space in an index, especially if the raw data has many unique terms that increase the size of the inverted index. Bloom filters, source metadata, and sourcetype metadata are much smaller in comparison and do not depend on the number of unique terms in the raw data.

According to the Splunk blog1, the best method for sizing or scaling a search head cluster is to estimate the maximum concurrent number of searches and divide by the number of CPU cores per search head. This gives you an idea of how many search heads you need to handle the peak search load without overloading the CPU resources. The other options are false because:

Estimating the maximum daily ingest volume in gigabytes and dividing by the number of CPU cores per search head is not a good method for sizing or scaling a search head cluster, as it does not account for the complexity and frequency of the searches. The ingest volume is more relevant for sizing or scaling the indexers, not the search heads2.

Estimating the total number of searches per day and dividing by the number of CPU cores available on the search heads is not a good method for sizing or scaling a search head cluster, as it does not account for the concurrency and duration of the searches. The total number of searches per day is an average metric that does not reflect the peak search load or the search performance2.

Dividing the number of indexers by three to achieve the correct number of search heads is not a good method for sizing or scaling a search head cluster, as it does not account for the search load or the search head capacity. The number of indexers is not directly proportional to the number of search heads, as different types of data and searches may require different amounts of resources2.

 The tailing_processor.log file would be the best place to search if you suspect there is a problem interpreting a regular expression in a monitor stanza. This log file contains information about how Splunk monitors files and directories, including any errors or warnings related to parsing the monitor stanza. The splunkd.log file contains general information about the Splunk daemon, but it may not have the specific details about the monitor stanza. The btool.log file contains information about the configuration files, but it does not log the runtime behavior of the monitor stanza. The metrics.log file contains information about the performance metrics of Splunk, but it does not log the event breaking issues. For more information, see About Splunk Enterprise logging in the Splunk documentation.

To resync a stale KV Store member in a search head cluster, you need to stop the search head that has the stale KV Store member, run the command splunk clean kvstore --local, and then restart the search head. This triggers the initial synchronization from other KV Store members12.

The command splunk resync kvstore [-source sourceId] is used to resync the entire KV Store cluster from one of the members, not a single member. This command can only be invoked from the node that is operating as search head cluster captain2.

The command splunk clean eventdata -local is used to delete all indexed data from a standalone indexer or a cluster peer node, not to resync the KV Store3.

According to the Splunk SmartStore Architecture Guide, the primary benefit of SmartStore is the separation of storage from compute resources within an indexer cluster. SmartStore enables Splunk to decouple indexer storage (data at rest) from the compute layer (indexers that perform searches and indexing).

With SmartStore, active (hot/warm) data remains on local disk for fast access, while older, less frequently searched (remote) data is stored in an external object storage system such as Amazon S3, Google Cloud Storage, or on-premises S3-compatible storage. This separation reduces the storage footprint on indexers, allowing organizations to scale compute and storage independently.

This architecture improves cost efficiency and scalability by:

Lowering on-premises storage costs using object storage for retention.

Enabling dynamic scaling of indexers without impacting total data availability.

Reducing replication overhead since SmartStore manages data objects efficiently.

SmartStore does not affect replication or search factors (Option A), does not handle Knowledge Object replication (Option C), and the Cluster Manager is still required (Option D) to coordinate cluster activities.

References (Splunk Enterprise Documentation):

• SmartStore Overview and Architecture Guide

• SmartStore Deployment and Configuration Manual

• Managing Storage and Compute Independence in Indexer Clusters

• Splunk Enterprise Capacity Planning – SmartStore Sizing Guidelines

The replication factor and the search factor are two important settings for a Splunk indexer cluster. The replication factor determines how many copies of each bucket are maintained across the set of peer nodes. The search factor determines how many searchable copies of each bucket are maintained. The default values for both settings are 3, which means that each bucket has three copies, and at least one of them is searchable

Splunk Enterprise performs a CRC check against the first and last 256 bytes of a file by default, as stated in the inputs.conf specification. This is controlled by the initCrcLength parameter, which can be changed if needed. The CRC check helps Splunk Enterprise to avoid re-indexing the same file twice, even if it is renamed or rotated, as long as the content does not change. However, this also means that Splunk Enterprise might miss some files that have the same CRC but different content, especially if they have identical headers. To avoid this, the crcSalt parameter can be used to add some extra information to the CRC calculation, such as the full file path or a custom string. This ensures that each file has a unique CRC and is indexed by Splunk Enterprise. You can read more about crcSalt and initCrcLength in the How log file rotation is handled documentation.

Search is not locked out when a customer has installed a 500GB Enterprise license and a 300GB, no enforcement license on the same license master. The no enforcement license allows the customer to exceed the license quota without locking search, but violations are still recorded. The customer can ingest up to 800GB of data per day without violating the license, but if they ingest more than that, they will incur a violation. However, the violation will not lock search, as the no enforcement license overrides the enforcement policy of the Enterprise license. For more information, see [No enforcement licenses] and [License violations] in the Splunk documentation.

Indexed extraction configurations are processed in the indexing phase of the Splunk Enterprise data pipeline. The data pipeline is the process that Splunk uses to ingest, parse, index, and search data. Indexed extraction configurations are settings that determine how Splunk extracts fields from data at index time, rather than at search time. Indexed extraction can improve search performance, but it also increases the size of the index. Indexed extraction configurations are applied in the indexing phase, which is the phase where Splunk writes the data and the .tsidx files to the index. The input phase is the phase where Splunk receives data from various sources and formats. The parsing phase is the phase where Splunk breaks the data into events, timestamps, and hosts. The search phase is the phase where Splunk executes search commands and returns results.

The curl https://serverhost:8089/services/admin/inputstatus/TailingProcessor:FileStatus command is used to check the status of the tailed files when troubleshooting monitor inputs. Monitor inputs are inputs that monitor files or directories for new data and send the data to Splunk for indexing. The TailingProcessor:FileStatus endpoint returns information about the files that are being monitored by the Tailing Processor, such as the file name, path, size, position, and status. The splunk cmd btool inputs list | tail command is used to list the inputs configurations from the inputs.conf file and pipe the output to the tail command. The splunk cmd btool check inputs layer command is used to check the inputs configurations for syntax errors and layering. The curl https://serverhost:8089/services/admin/inputstatus/TailingProcessor:Tailstatus command does not exist, and it is not a valid endpoint.

According to the Splunk documentation1, in a search head cluster, the KV Store Primary is the same node as the search head cluster captain. The KV Store Primary is responsible for coordinating the replication of KV Store data across the cluster members. When any node receives a write request, the KV Store delegates the write to the KV Store Primary. The KV Store keeps the reads local, however. This ensures that the KV Store data is consistent and available across the cluster.

 ITSI can impact the planning of a Splunk deployment depending on the Key Performance Indicators (KPIs) that are being tracked. KPIs are metrics that measure the health and performance of IT services and business processes. ITSI collects, analyzes, and displays KPI data from various data sources in Splunk. Depending on the number, frequency, and complexity of the KPIs, additional infrastructure may be needed to support the data ingestion, processing, and visualization. ITSI does not require a dedicated deployment server, nor does it affect the number of users using ITSI. ITSI in a Splunk deployment does require additional hardware resources, such as CPU, memory, and disk space, to run the ITSI components and apps

KV Store is a feature of Splunk Enterprise that allows apps to store and retrieve data within the context of an app1.

KV Store resides on search heads and replicates data across the members of a search head cluster1.

KV Store resiliency refers to the ability of KV Store to maintain data availability and consistency in the event of failures or disruptions2.

One of the factors that affects KV Store resiliency is the network latency between search heads, which can impact the speed and reliability of data replication2.

Decreasing latency between search heads can improve KV Store resiliency by reducing the chances of data loss, inconsistency, or corruption2.

The other options are not directly related to KV Store resiliency. Faster storage, indexer CPU and memory, and Operations Log size may affect other aspects of Splunk performance, but not KV Store345.

According to the Splunk documentation1, the CHARSET setting in props.conf specifies the character set encoding of the source data. This setting has the least impact on indexing performance, as it only affects how Splunk interprets the bytes of the data, not how it processes or transforms the data. The other options are false because:

The SHOULD_LINEMERGE setting in props.conf determines whether Splunk breaks events based on timestamps or newlines. This setting has a significant impact on indexing performance, as it affects how Splunk parses the data and identifies the boundaries of the events2.

The TRUNCATE setting in props.conf specifies the maximum number of characters that Splunk indexes from a single line of a file. This setting has a moderate impact on indexing performance, as it affects how much data Splunk reads and writes to the index3.

The TIME_PREFIX setting in props.conf specifies the prefix that directly precedes the timestamp in the event data. This setting has a moderate impact on indexing performance, as it affects how Splunk extracts the timestamp and assigns it to the event

 The following is a good practice for a search head cluster deployer: The deployer must be used to distribute non-replicable configurations to search head cluster members. Non-replicable configurations are the configurations that are not replicated by the search factor, such as the apps and the server.conf settings. The deployer is the Splunk server role that distributes these configurations to the search head cluster members, ensuring that they have the same configuration. The deployer does not only distribute configurations to search head cluster members when they “phone home”, as this would cause configuration inconsistencies and delays. The deployer does not distribute configurations to search head cluster members to be valid configurations, as this implies that the configurations are invalid without the deployer. The deployer does not only distribute configurations to search head cluster members with splunk apply shcluster-bundle, as this would require manual intervention by the administrator. For more information, see Use the deployer to distribute apps and configuration updates in the Splunk documentation.