Lima Backbone extracts every table that holds data from the ECC, Oracle EBS, PeopleSoft or BW system you are retiring — at terabyte scale, with record counts proved for each table — lands it under retention rules, keeps it queryable for auditors and analysts through the fabric, and restores any dataset into any target when a process, an audit or a dispute needs it. The licence, hardware and support contracts can end.
A system can be switched off once its data is complete somewhere else, readable for the retention period, answerable to auditors, erasable on request, and restorable when a dispute or a process needs it. Each obligation maps to something Lima already does at volume.
Full-table extraction in parallel, with record counts and control totals compared against the source for each table. The report is the evidence that nothing was left behind.
History lands as Parquet tables in your lakehouse or object storage with the catalog, lineage and access policy applied, so it stays readable without Lima and governed with it.
Auditors and analysts query the archive through ChatLima and SQL using the same vocabulary as before. Every access is logged.
Erasure runs against the archived records with evidence attached — the 47-minute erasure workflow from the showcase estate applies to archived data too.
The same engine reloads a dataset into S/4HANA, a warehouse or a file target, reconciled the same way it was extracted.
Each week has a deliverable, and finance and audit sign the last one.
Put your own figures in. The result is arithmetic on what you enter.
SAP's own archiving and lifecycle tools can hold the compliance status your auditors expect; Lima holds the same history in a form people can query. Where a vault already exists, Lima feeds it and reads from it.
The finance team asks the archive the same questions it asked the ERP, through ChatLima or SQL, and gets answers with the access policy applied.
The reconciliation report — counts and control totals for every table — is what auditors accept as evidence of a complete extraction.
Cold tiers for old records, warm tiers for what is still consulted; cost follows the retention class.
ECC, Oracle EBS, PeopleSoft, BW and custom applications retire through the same six-week pattern.
Retirement specialists report 30–70% smaller migration volumes when history is retired ahead of the S/4HANA move.
Reloading a dataset uses the same parallel loads and reconciliation as the extraction, into whichever target asks for it.
One engine underneath every page in this family. It reads from any source and writes to any target, runs batch loads or change-data streams, handles a nightly delta or a multi-terabyte migration, and is licensed by capacity, so the bill stays the same however many rows move. Lima's fabric answers questions on data in place; Lima Backbone moves, transforms and returns data on a deadline. When to federate and when to pipeline · The Backbone hub.
For as long as each retention class requires, on the tier that class specifies, in your own lakehouse or object storage. Retention schedules and legal holds are rules in Lima that run on a schedule and leave an audit trail.
Erasure requests run against the archived records with the evidence attached, using the same workflow that handles live data. Legal holds take precedence where they apply.
They get a read-only role with logged access, and they can use ChatLima, SQL or the export they ask for. The archive is also readable in Parquet by any query tool your audit firm brings.
Yes. Restore reloads any dataset into S/4HANA, a warehouse or a file target with the same reconciliation. That is also how a legal or tax request is served without keeping the old system alive.
The Terabyte Run — a multi-terabyte, billion-plus-row workload covering a full-table archive extract, a reconciled migration and a restore — is scheduled; the methodology is published first and the results follow. The largest documented run today is about 220 million rows in roughly 30 minutes (estimate).
SAP ECC and BW, Oracle EBS, PeopleSoft and custom applications with a relational database. The inventory step tells you the table count and volume before the fixed price is set.
Fixed price per system: inventory in week 1, extraction and reconciliation by week 6, switch-off criteria signed by finance and audit.
