Give Cloud Apps Live Access to Mainframe VSAM Data — Without Copying It

Every plan to use mainframe data in the cloud starts the same way: "first, make a copy."

In this Skyward Data episode, VirtualZ Computing Founder & CEO Jeanne Glass and Co-Founder & CTO Vince Re — the original author of CA Top Secret — explain how to give cloud, distributed, and AI applications live, real-time read and write access to mainframe VSAM data without moving or copying it, using VirtualZ Lozen. 

Why does every mainframe-to-cloud project reach for a copy first?

Because mainframe data is complex — VSAM, EBCDIC, mainframe-specific formats — so the reflex is to extract it, transform it, and replicate it to a data lake, warehouse, or database. But every copy is a duplicate to store, secure, and govern; a pipeline to build and maintain; and a source of drift from the live system of record.

Vince shares how one financial-services company spent two years trying to make a change-data-capture (CDC) model work in production before giving up. Lozen takes the copy out of the middle. The data never leaves z/OS. Applications running in AWS, Azure, Google Cloud, or any distributed environment read and write VSAM, QSAM, sequential files, PDS, and GDG data using standard file-system calls — the mainframe presented as a standard NFS network drive, with no proprietary API and no application rewrite.

Because there's only one copy of the data, you get real-time integrity on reads and writes: no stale data, no cross-platform locking, no reconciling updates back to the mainframe. Existing RACF, ACF2, and Top Secret security still governs every access — even from a remote cloud app — and the work is zIIP-eligible to contain MIPS. 

Jeanne and Vince also cover a problem most teams miss: because mainframe data is shared data, moving one application off the mainframe usually forces you to move ten others that share its files. Lozen ends that — modernize one app and leave its data where it is. And they draw an honest line between live access and replication: use Lozen for real-time, file-system read/write access where the mainframe stays authoritative; use PropelZ (VirtualZ's no-code replication product) for SQL/relational access, tools like Tableau or Snowflake, or heavy AI query offload. Most large mainframe sites need both. 

In this episode:

• Why "just make a copy" is the reflex for mainframe data — and what it really costs

• How Lozen gives cloud and distributed apps live read/write access to mainframe VSAM data with no movement

• Real-time data integrity for reads and writes, with no locking or reconciliation headaches

• How existing RACF/ACF2/Top Secret security is preserved and how Lozen is zIIP-eligible

• The shared-data problem: modernize one app without moving the ten that share its files

• Live access vs. replication — when to use Lozen and when to use PropelZ 

If you're a mainframe architect, data platform leader, application modernization or migration lead, or an AI and analytics owner getting IBM Z data into the cloud, this conversation is for you. 

See Lozen in action — including the AWS PartnerCast featuring Lozen with Rocket Enterprise Server and AWS services — and explore more at https://virtualzcomputing.com/lozen. Skyward Data is the podcast from

VirtualZ Computing — the first woman-founded, woman-led, and Native American-founded and -led mainframe software provider. VirtualZ's no-code, install-and-run portfolio (PropelZ, FlowZ, Lozen, and Zaac) connects IBM Z mainframe data to cloud, AI, and analytics without rewrites, pipelines, or forced migrations. 

Learn more: https://virtualzcomputing.com