What is our primary use case?
The main use cases for Splunk User Behavior Analytics include threat detection. I detect insider threats, compromised users, account misuse, and all those things. I use unsupervised and supervised ML models. The risk scoring is another feature I use with categories. I assign risk scores to users and entities. I also do the integration with Splunk because Splunk User Behavior Analytics natively integrates with Splunk Enterprise. I build timeline visualizations, threat timelines, and event linkage. I also perform alert prioritization and threat-related prioritization based on trends.
What is most valuable?
The best features in Splunk User Behavior Analytics include anomaly detection, behavioral profiling, and risk scoring and prioritization functionality. There are certain out-of-the-box use cases as well, such as insider threat detection and credential misuse detection. The solution also includes ML-related features.
I have been using customizable dashboards and reports in Splunk User Behavior Analytics, and it has built-in dashboards that show top users by risk, trending anomalies, and additional metrics.
I evaluate the automation capabilities for threat detection in Splunk User Behavior Analytics, which uses automated machine learning models and behavioral analytics to detect complex and hidden threats. For example, unsupervised ML models can learn normal behavior over time such as logon patterns and access levels. The system can detect deviations in user behavior such as logins during odd hours or from new locations.
What needs improvement?
There are improvements that could be made to Splunk User Behavior Analytics as any product will have advantages and disadvantages. Scalability is one consideration. For example, the advantages include rapid auto scaling to meet demand. A disadvantage is that it can lead to cost overrun if not properly factored or governed. The speed of deployment offers faster provisioning as an advantage, but it can require substantial automation skills and infrastructure as code expertise, which can be challenging.
Cloud provides major operational benefits such as agility, automation, resilience, and global access when setting up on Cloud. However, it introduces challenges such as cost control, complexity, and vendor dependency. For example, global reach allows deployment of apps and services closer to users worldwide, but data sovereignty concerns exist and region selection must align with compliance requirements.
For how long have I used the solution?
I have been dealing with Splunk User Behavior Analytics for one year.
What do I think about the stability of the solution?
The solution is generally stable. Although no Cloud deployment is 100% stable, the platform uptime SLA on the Cloud is 99.9%. The stability depends on how infrastructure redundancy is factored into the design and implementation. With built-in redundancy across zones and regions, 99.9% uptime is achievable. Automated scaling must be enabled to maintain better platform uptime.
How are customer service and support?
Splunk offers multiple types of support tiers including standard, premium, and enterprise support. From the responsiveness perspective, Splunk is very responsive with SLA-bound support for premium tiers. There are numerous active forums, user groups, and Splunk Answers as part of their community support.
The knowledge base includes extensive articles, thorough documentation, and Splunk-based apps are available. The support quality is excellent for paid tiers, following enterprise-grade SLAs with proactive support and deep expertise.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
I have worked with similar solutions from other vendors, including Elasticsearch's product called Elastic. There are crucial differences between Elastic and Splunk User Behavior Analytics, with Splunk having its own advantages. Splunk has proprietary log analytics capabilities, while Elastic is built on an open source base.
For example, when working with Citrix UberAgent, an end-user experience monitoring tool, it has extensive integration with Splunk, allowing creation of 13-14 dashboards. Elastic has limitations, enabling only three to four dashboards with Citrix. Regarding licensing models, Splunk uses commercial volume-based pricing, while Elastic offers an open source basic version with platinum and enterprise tiers for full features. This provides an advantage for smaller organizations that don't require extensive features.
How was the initial setup?
The initial setup and deployment requires several key steps. Requirement gathering involves defining business needs, application architecture, compliance, and security requirements. Cloud platform selection must be made between Azure, AWS, GCP, or hybrid/multi-Cloud options.
Design architecture involves solutions based on computing, storage, network security. Subscription account setup requires creating Cloud accounts, defining billing structure, and cost centers. Identity and Access Management (IAM) setup includes configuring AD, Azure AD, or AWS IAM, defining roles, policies, and RBAC.
Network setup involves configuring virtual networks, subnets, NSC, firewalls, VPN, or express route. Core infrastructure provisioning includes creating virtual machines, Kubernetes clusters, databases, and storage. Security configuration requires implementing encryption, endpoint protection, and firewall security groups.
Additional steps include compliance checks, application deployment through CI pipelines or Azure DevOps, monitoring setup using tools such as Azure Monitor, AWS CloudWatch, and Splunk. Backup configuration, performance tuning, user acceptance testing, go-live procedures, and ongoing optimizations are also necessary components of the setup process.
What's my experience with pricing, setup cost, and licensing?
In terms of setup cost, pricing, and licensing, Splunk User Behavior Analytics is not an inexpensive product. The setup requires numerous components including storage, networking, identity access, integration, and backup. Cost-wise, with pay-as-you-go options, the pricing can be extensive.
The licensing model includes various options. Reserved instances with one or three-year commitments offer lower rates, providing up to 70% savings. The Bring Your Own License (BYOL) option allows use of existing Microsoft, Red Hat, or Oracle licenses with software assurance. The Service Provider License Agreement (SPLA) model offers monthly usage-based licensing for service providers.
What other advice do I have?
I am currently using Splunk User Behavior Analytics on Cloud in a SaaS model. The customer purchased the solution directly, not through AWS. We operate as a service provider managing the solution. I have rated Splunk User Behavior Analytics 9 out of 10.
Which deployment model are you using for this solution?
Public Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Amazon Web Services (AWS)