SAP System & Application-Server Architecture on AWS

Your AAS sit in two AZs for HA and some critical batch jobs run slow. What architecture choice fixes it, and why isn't even-spreading the goal here?

On AWS, is an 'SAP instance' the same thing as an 'EC2 instance', and what one property distinguishes them?

On the same instance (e.g. x2iedn.32xlarge), TDI v5 allows 4 TiB for HANA vs 2.7 TiB under TDI v4. What's the catch that decides which spec you can actually claim?

Choosing storage for SAP app-server volumes vs shared file systems on AWS — what's the discriminating factor, and why isn't the certified-but-pricey tier the default?

For DR of the SAP database itself, why reach for database-native replication instead of (or alongside) storage-based DRS?

In an SAP HA setup, which component owns the SAP lock table and which keeps it from being lost on failover — and why must ERS sit where it does?

Sketching multi-AZ HA for SAP NetWeaver on AWS, what are the load-bearing pieces — and which one do people forget, breaking failover routing?

A customer wants to lift-and-shift their on-prem SAP HA stack to AWS. Which two HA solutions must you tell them are NOT supported, and what are the supported alternatives?

A distributed (multi-host) SAP deployment splits DB and ASCS+PAS onto separate EC2 instances in one AZ. Does that buy HA, and when is it still the right call?

An ABAP developer asks whether to write Open SQL or native SQL for code that must run on any supported database. Which, and what does the database interface actually do?

You want to cut cost on SAP app-server compute. Can you put dev/sandbox app servers on Graviton, and how do AMD vs Intel split otherwise?

You're designing HA for SAP Cloud Connector on AWS. Do you wrap it in an OS cluster like ASCS, and how does its failover actually work?

PAS vs AAS — beyond 'first one vs the extras', what property actually matters for HA design, and what's the common misconception about their capability?

For SAP support, which SAP components must live on one cloud provider — and does an RFC link to an on-prem system break support?

Why can't you use a normal VPC IP (or Elastic IP) for SAP HA failover across AZs on AWS, and what makes an overlay IP reachable from on-prem?

Using AMI-copy for SAP app-server DR, what event forces you to recreate the AMI, and what's the cheapest way to cut RTO below a from-scratch boot?

How many overlay IPs does a single clustered SAP system burn from your RFC 1918 range, and why does that number force a planning tool at scale?

Of the three SAP tiers, which two are bound by the single-cloud-provider support rule, and which one is exempt — and why?

SAP Fiori vs SAP GUI behind an AWS load balancer — which ELB type for each, and what single property of the traffic decides it?

Making SAP Content Server highly available — why isn't load-balancing two app instances enough, and what HA pattern does it actually need?

A single-host SAP deployment puts ASCS, PAS and the DB on one EC2 instance. What SLA are you bound to, and where's the line below which it's acceptable?

A reviewer asks for the 'SAP certification' covering AWS Elastic Disaster Recovery (DRS) for your SAP DR. What do you tell them, and who's the support contact?

Three ways to make SAP Web Dispatcher HA on AWS differ in what failure they survive. Which one covers only software faults, and which is preferred?

ASCS failover on AWS can be manual-with-alerts or automated-via-cluster. Which one protects you from losing SAP locks, and what's the manual path's hidden work?

You want Amazon RDS for SAP BusinessObjects' CMS/audit databases. What constraint does that put on where the application tiers run?

You're running SAP on an Oracle database on AWS. What OS must the SAP application servers use, and which SAP Note pins this?

Someone says the PAS isn't a SPOF because its work processes are fault tolerant. Where does that argument break, and which protection gives zero downtime?

Across cold / warm-hot / AWS DRS for SAP app-server DR, which option gives low RPO/RTO at only medium cost — and why does it fit app servers specifically?

When designing a reliable SAP system on AWS, which parts of resiliency are yours to own and which does the cloud provider own?

You move SAP S/4HANA to AWS EC2 and someone calls it 'the cloud version of SAP'. Why is that wrong, and what does the deployment model actually give you?

An architect wants to put SAP application servers behind EC2 auto-scaling to handle load spikes and save cost. What's the blocker?

For SAP shared storage and load balancing on AWS you *can* run NFS servers and a third-party LB on EC2. Why is that the wrong default now?

Why does AWS have to get its platforms 'certified' by SAP at all, and which single component of that certification carries the most weight?

You're asked to prove an SAP config is fully supported on AWS. From which angles do you check it, and which SAP Notes are the authority?

Your SAP landscape includes a non-SAP archiving tool that integrates with SAP ERP. What must you verify before migrating it to AWS, and how should you schedule it?

Even in a fully HA, fault-tolerant SAP setup on AWS, some failover still isn't automatic 'for free'. Which components still need application-layer failover management?

Your users are global, so the instinct is to spread one SAP landscape across multiple AWS regions. Why does that usually not work, and what still matters?

If most of an SAP system has to run on EC2, in what sense can you still 'go serverless' on AWS, and why bother?

Before designing HA, an architect needs the full inventory of what can fail in an SAP system. What are the single points of failure to protect?

People say 'the SAP production system' as if a product were one system. How is an SAP product actually structured across environments?

Which Linux distributions are certified to run SAP application servers on AWS, and what constrains OS version choice within them?

You can run plain RHEL or SLES for SAP on AWS — why does the book push the 'for SAP' flavors instead, and what do you actually gain?

What role does SAP Cloud Connector play in an SAP-on-AWS landscape, and how does it differ architecturally from a NetWeaver app server?

Two SAP systems store identical backup sizes in S3 but one restores far slower. Given the same backup tooling, what instance property explains the gap in RTO?

Rather than hand-picking every EC2 type, EBS config and VPC setting for a new SAP install, what AWS-native service streamlines it and what does it guarantee?