Memory, not megawatts.
A HANA host is full long before its power is: the operating system takes its share, every system reserves an allocation limit, and replicas wait for a takeover. Rymvard counts all of it, host by host.
New · SAP and HANA hosting
Every HANA gigabyte.Every takeover, checked.
For SAP hosting providers and SAP operations teams: Rymvard shows the memory of every HANA host, checks every system-replication takeover and plans room for the next go-live — in the same ledger as power, racks and compute.
Read-only. The screens show a simulated landscape of 200 HANA hosts — not a customer.
Memory, not megawatts, decides.
An SAP hosting estate rarely runs out of power first. It runs out of HANA memory — on the wrong host, at the wrong site, in the quarter a customer goes live. And the one moment that matters most, a takeover, is the one nobody has tried.
A takeover nobody has tried.
System replication looks healthy until the day it is needed — and the secondary host turns out to be filled with test systems. Rymvard checks every replicated system before that day.
The 2027 wave.
SAP ECC leaves mainstream maintenance at the end of 2027. Every move to S/4HANA needs HANA memory, often twice for a while. Rymvard sets planned go-lives against the memory that is really free.
What Rymvard shows
One view of every HANA host.
The same numbers for SAP basis, sales and the board — each traceable to the host, the system and the moment it was observed.
Memory per host.
Operating-system reserve, used (thirty-day peak), allocated but unused, held for replicas and free — for every HANA host, every site and the whole estate. Allocations beyond physical memory are shown, never hidden.
A takeover check per system.
Can the secondary host carry production after a takeover, once the test systems that may stop have stopped? Fits, fits after stopping, does not fit — or unknown when a figure is stale.
Placement and runway.
A new system gets the host with the best fit and a partner host for its replica. Planned go-lives, weighted by their probability, become a memory runway per site and quarter.
Customers kept apart.
System IDs repeat across customers. Rymvard keys every system by ID, site and customer, shows each customer's share and mirrors HANA into the compute view for showback and lifecycle.
Read-only, from the systems you run.
Rymvard never writes to SAP. A small collector at each site reads what your landscape already knows and keeps it current.
- SAP Host Agent — instances and hosts, over HTTPS, with read operations only; the collector refuses anything else.
- HANA monitoring views — memory, replication and tenant databases from a short, fixed list of read-only queries, with a monitoring user over TLS.
- Landscape tools — SAP Focused Run, ServiceNow, SAP Solution Manager and SAP LaMa — for systems, customers and environments.
- Detailed records stay in their country; customer names are visible only to the people who need them.
Workspace · SAP Operations
A home for SAP basis.
The SAP Operations workspace opens on four figures and the findings that concern SAP — with AI that explains every number and suggests the next step. See the workspaces
- HANA memory freeWhat can the next system still take, after reserves and replicas?
- Production without replicationWhich production systems would a site loss take down?
- Takeovers that would failWhere would the secondary host run out of memory?
- HANA hosts over 90 %Which hosts are close to their limit at peak?
Europe and Asia-Pacific
Four SAP hosting scenarios.
The S/4HANA wave in Frankfurt, a DORA resilience test in Luxembourg, scarce power in Singapore and in-country SAP in Mumbai.
Frankfurt · SAP hosting provider
Make room for the S/4HANA wave
SAP ECC leaves mainstream maintenance at the end of 2027. Every move to S/4HANA needs HANA memory — for a while, twice.
Luxembourg · Managed SAP for financial services
Prove that every SAP takeover will hold
Under DORA, banks must show that critical systems recover. For SAP that means a takeover the replica host can really carry.
Singapore · Regional SAP hosting hub
Find the HANA memory you already own
New data-center power is allocated in selective calls. The fastest new HANA capacity is memory that is reserved but never used.
Mumbai · In-country SAP hosting
Grow SAP inside the country
Payment data must stay in India. Every new SAP landscape needs in-country HANA capacity — and a plan for it.
Sources for the SAP and HANA scenarios: SAP, maintenance for SAP Business Suite 7 including ECC 6.0 — mainstream until the end of 2027, optional extended maintenance until the end of 2030 (announced 4 February 2020); Regulation (EU) 2022/2554 on digital operational resilience (DORA), applicable from 17 January 2025; Monetary Authority of Singapore, Notice 644 on Technology Risk Management — a recovery time objective of no more than four hours for each critical system; IMDA and EDB, Data Centre Call for Application (about 80 MW allocated in 2023, a second call from December 2025); Reserve Bank of India, Storage of Payment System Data (circular of 6 April 2018). Scenarios are illustrative; no customer, site or outcome is implied.
Start with one landscape.
We map your HANA hosts and systems and show what is really free, reserved and at risk.