BlazeMeter 3.5 automatically splits CSV test data across browser-test engines, giving release owners unique user datasets without manual file preparation. The change reduces a common source of invalid multi-engine results: repeated test identities, accounts, or transactions assigned to more than one engine.
Unique test data for browser-based testing
Scripted and scriptless browser-based tests can now distribute splittable CSV data automatically across engines. Each engine receives a distinct, non-overlapping subset, matching the established Split CSV behavior for standard performance tests.
For teams operating distributed browser tests, this changes the preparation workflow:
- Remove manual pre-splitting of source CSV files.
- Reduce the risk of concurrent engines using the same credentials or records.
- Preserve per-engine CSV files in engine artifacts for investigation.
- Enable the Split CSV option in the scriptless editor when eligible data is present.
This is particularly relevant where test data represents customer-like records, inventory, payment tokens, or other finite datasets. Unique allocation improves confidence that observed failures reflect application behavior rather than test-data collisions. Artifact retention also gives QA leads a practical audit trail when reviewing a failed run.
Faster API Monitoring for critical services
API Monitoring now offers an Every 30 seconds schedule in a monitor's Daily schedule on eligible paid plans. This is the first supported frequency below one minute; existing schedules from one minute to one day are unchanged.
A 30-second interval can shorten the time between an endpoint failure and its first observed monitoring signal. It should be applied selectively, not as a default for every monitor. Higher-frequency execution increases run volume, alert traffic, and the need for disciplined threshold settings.
Good candidates include:
- Authentication and token-issuance endpoints.
- Checkout, order-submission, and payment dependencies.
- APIs with strict availability commitments.
- Services that act as release gates during deployment windows.
Release and operations leaders should document which endpoints warrant this interval, define escalation ownership, and review whether alert routing can support the faster detection cadence.
Taurus Cloud vulnerability remediation
The current Taurus Cloud version, 2.4.523-reduced, resolves identified high-severity vulnerabilities by applying required fixes and removing affected components. The release notes identify:
- CVE-2026-69192 in the Node
ip-addresscomponent. - CVE-2026-73566 in the Node
tarcomponent.
For security and platform teams, the relevant action is not merely updating a tool record. Confirm the deployed version, capture the remediation status in vulnerability tracking, and retain release-note evidence alongside the approved change. This makes the control easier to demonstrate during internal review or external audit.
Bounded memory configuration for virtual services
Service Virtualization now permits an account-level maximum Java heap limit for overrides and per-virtual-service caps on private locations. Teams can still tune individual services, but only within an approved memory boundary.
This helps reduce operational uncertainty when several virtual services share private-location capacity. Without defined limits, a local adjustment for one service can affect the available resources and stability of neighboring test workloads.
A practical governance model is to:
- Set an account maximum based on private-location capacity.
- Define lower service caps according to expected traffic and payload size.
- Require review for exceptions or higher-memory services.
- Monitor heap use and failed starts after configuration changes.
The controls make capacity planning more predictable, especially where performance testing and service virtualization run in the same managed environment.
Release-management implications
BlazeMeter 3.5 combines workflow improvements with controls that matter to enterprise delivery governance. Test-data distribution addresses result validity. Faster monitoring can reduce detection delay. Security remediation supports vulnerability closure. Memory caps establish clearer resource accountability.
Teams should update runbooks and release checklists to reflect the new options. A small pilot with representative browser tests, priority API monitors, and private-location virtual services will expose configuration assumptions before wider adoption.
How Merito helps
Merito helps organizations assess BlazeMeter 3.5 against existing QA, DevOps, and security controls. We can define adoption guardrails, configure monitoring and test-data practices, validate private-location capacity, and produce evidence suitable for release and audit reviews.
Merito is an authorized Perforce reseller and services partner. Our Perforce BlazeMeter team can scope licensing, sizing, and rollout for 3.5, and our enterprise upgrade services help you plan and validate the upgrade with minimal disruption to release schedules.