Explanation:

Why B is Correct?

Tableau Server generates detailed application logs that capture errors, warnings, and crashes specific to Tableau processes.

These logs (e.g., tabadmin.log, vizqlserver.log, backgrounder.log) help identify Tableau-specific issues (e.g., service failures, memory leaks, or configuration errors).

Since the issue is intermittent crashes, Tableau logs are the first place to check for root-cause analysis.

Why Other Options Are Less Relevant?

A. Windows Event Viewer System Logs are useful for OS-level issues (e.g., hardware failures, driver crashes) but are secondary if Tableau itself is crashing.

C. IIS Logs are only applicable if Tableau is configured with IIS (default is embedded Apache/Nginx). Even then, IIS logs track web requests, not application crashes.

D. Network Logs would help with connectivity issues but are unrelated to a server crash.

Reference:

Tableau’s official documentation on Server Log Files recommends checking logs in the C:\ProgramData\Tableau\Tableau Server\data\tabsvc\logs directory for troubleshooting crashes.

Next Steps:

Check tabadmin.log for high-level server errors.

Review vizqlserver.log for rendering-related crashes.

Look for OutOfMemory errors in backgrounder.log if background tasks are failing.

Explanation:

Why Option B is Correct:

Cache optimization in Tableau Server balances performance (reduced query load) with data freshness.

Clearing the cache on a schedule (e.g., after ETL jobs refresh source data) ensures:

Performance benefits of cached queries.

Data consistency by invalidating stale cache entries.

Reference: Tableau Server Performance Tuning Guide.

Why Other Options Are Incorrect:

A) Refresh cache on every view load:
Defeats the purpose of caching (no performance gain).

C) Disabling caching:
Forces real-time queries, increasing server load and latency.

D) Increasing RAM:
Helps caching but doesn’t address cache invalidation logic.

Key Cache Settings in Tableau Server:

vizqlserver.session.expiry.timeout: Controls how long cached sessions persist.

Backgrounder tasks: Schedule cache clears post-data-refresh.

Explanation:

Why Option B is Correct:

High load on the primary node indicates that the current setup cannot handle the workload efficiently.

Adding worker nodes allows Tableau Server to

Distribute processes (e.g., VizQL, Backgrounder) across multiple machines.

Improve scalability and fault tolerance.

Reduce bottlenecks on the primary node.

Reference: Tableau Server Scalability Guide.

Why Other Options Are Incorrect:

A) Switching OS:

Unlikely to resolve performance issues caused by hardware/resource limits.

C) Increasing bandwidth:

Helps with network congestion but not CPU/memory bottlenecks on the primary node.

D) Consolidating processes:

Worsens the problem by overloading the primary node further.

Steps to Scale with Worker Nodes:

Assess load: Use Tableau Server Admin Views to identify which processes (VizQL, Backgrounder) are overloaded.

Add nodes: Use tsm topology commands to deploy additional worker nodes.

Reassign processes: Balance VizQL/Backgrounder across nodes.

Explanation

In large, high-security, high-performance Tableau Server deployments (especially for financial institutions), the service-to-node topology should be strategic:

Critical services (e.g., Repository [pgsql], Data Server, Coordination Service) handle sensitive data and are central to server operations. These should be:

. Isolated on dedicated nodes to prevent resource contention and limit exposure.
. Hardened with security configurations, limited access, and strong OS/network protections.

Less critical or stateless services (e.g., VizQL Server, Application Server) can be collocated on shared nodes to optimize resource usage without jeopardizing security.

This approach balances security (minimizing the attack surface for critical components) and performance (avoiding CPU/memory contention).

Why not the others?

A. Isolating all services on individual nodes → Overly complex, costly, and often unnecessary. It doesn’t give a meaningful performance/security boost for non-critical services.

B. Collocating all services on a single node → Creates a single point of failure and resource bottlenecks, not suitable for high-security environments.

D. Randomly distributing services → No control over performance or security; risks inefficiency and possible downtime.

Reference:

Tableau Help: Distributed and High Availability Deployments
Tableau Blueprint: "Isolate critical Tableau Server processes on dedicated hardware to enhance both performance and security in large-scale deployments."

Explanation:

Why B is Correct?

A phased migration plan minimizes disruptions by:

Testing small batches of dashboards/data sources first to catch issues early.

Scheduling rollouts during low-usage periods (e.g., weekends).

Validating functionality (e.g., permissions, subscriptions) post-migration.

Tableau’s Migration Guide emphasizes this approach.

Why Other Options Are Incorrect?

A. Speed over planning: Risks broken dashboards, data loss, or downtime.

C. Migrating largest datasets first: May overwhelm the new Server and delay critical fixes.

D. Ignoring data/dashboards: Defeats the purpose of migration—users need their content!

Key Steps for a Smooth Migration:

Inventory content:

List dashboards, data sources, users, and schedules.

Prioritize by business impact:

Migrate mission-critical content first.

Test in staging:

Validate performance and permissions.

Reference:
Tableau’s Pre-Migration Checklist.

Final Note:

B is the only method ensuring continuity. Options A/C/D risk chaos. Always communicate timelines to users.