Evidence & Validation
What Others Are Saying
See how VirtualZ software performs in real enterprise environments, and what customers, technology partners, and independent industry sources say about it.
By leveraging PropelZ for Linux within our banking platform strategy, we’ve increased throughput, reduced unnecessary processing layers, and improved operational efficiency all without introducing additional complexity into our systems.
We're seeing PropelZ sustain around 56,000 records per second, and in one run it moved about 60.9 million records in 18 minutes. Across 100 to 200-million-record datasets, that's roughly twice the throughput of the managed file transfer-based approach — and the data comes through intact, with no impact to production, whether it's VSAM, non-VSAM, tape, or GDG.
Lozen was incredibly easy to install and configure. The step-by-step documentation made the process seamless, and with responsive support, everything worked as expected on the first try. Even with our complex AWS and mainframe environment, Lozen integrated smoothly, ensuring secure, real-time data access.
VirtualZ is solving complex data access challenges with an innovative approach, and we’re proud to support that with the performance and reliability of our DataDirect connectivity. Together, we’re helping customers simplify integration and bring legacy data into modern environments faster.
PropelZ sustained approximately 56,000 records per second across 100–200 million-record datasets. In one observed run, 60.9 million records moved in 18 minutes, roughly twice the throughput of the customer’s managed file transfer approach, with no observed production impact.
- Dataset size
- 100 to 200 million records
- Data types
- VSAM, non-VSAM, tape, GDG
- Single run
- 60.9 million records in 18 minutes
- Comparison
- Roughly twice the customer's managed file transfer approach
- Production impact
- None observed
460,000 VSAM records were loaded to PostgreSQL on Google Cloud in 25 seconds.
- Workload
- 460,000 VSAM records
- Target
- PostgreSQL
- Platform
- Google Cloud
- Result
- 25 seconds
On the same ~400,000-record file and load, processing time dropped from about two hours to under five minutes after switching to batch commit.
- Workload
- ~400,000 records
- Before
- Approximately two hours
- After
- Under five minutes
- Change
- Batch commit
- Qualifier
- Configuration/tuning gain, not an out-of-the-box product baseline
PropelZ delivered higher throughput for mainframe-to-cloud and database movements (~500–600 rows/sec to ~6,000 rows/sec, 8–10x improvement) using an optimized Progress DataDirect JDBC driver.
- Improvement
- From ~500–600 rows/sec to ~6,000 rows/sec, an 8–10x improvement
- Comparison
- Open-source JDBC drivers
- Driver
- Progress DataDirect JDBC
Approximately 9,000 mixed-size tape files were processed in about 17.5 hours with 100 transfers running in parallel. Concurrency was deliberately capped at 100 to avoid impacting other LPAR workloads.
- File set
- ~9,000 mixed-size tape files
- Runtime
- ~17.5 hours
- Parallelism
- 100 concurrent transfers
- Constraint
- Concurrency capped to protect other LPAR workloads
- Qualifier
- Constrained run, not a maximum rate
PropelZ sustained about 26,000 64 KB records per second while consuming approximately 3% of a single zIIP engine, moving multiple terabytes per hour.
- Sustained rate
- ~26,000 records per second at 64 KB
- zIIP utilization
- Approximately 3% of a single engine
- Volume
- Multiple terabytes per hour
Explore the evidence
Questions about these results
Can VirtualZ software handle enterprise data volumes?
Yes. PropelZ has sustained approximately 56,000 records per second, roughly six terabytes an hour, across datasets of 100 million to more than 200 million records. It streams directly to cloud storage to remove the file-size limits that staging creates. For live access, a global systems integrator chose Lozen for VSAM files shared by hundreds of applications, where custom approaches could not work.
Does VirtualZ software disrupt production workloads?
How does VirtualZ software make mainframe data usable for AI and analytics?
How quickly can VirtualZ software be installed and running?
Does VirtualZ software fit enterprise mainframe security and operations?
How do customers, outsourcers, and systems integrators validate VirtualZ software?
Your story starts with one dataset.
Bring us the workload that matters most to your business. In a 30-minute briefing, we’ll show you how VirtualZ can connect your mainframe data to the cloud, AI, and analytics tools you already use — in your own environment.