SAP HANA Architecture & Persistence

Column-store writes are costly — how does SAP HANA mitigate that?

When reading a column table, where does HANA look besides the main column store?

What role do indexes still play in SAP HANA column tables?

What's the difference between the SAP HANA nameserver and indexserver?

What's the SAP HANA service port schema for instance number xx?

Where are SAP HANA database parameters stored, compared with the instance profile directory?

In what respects are SAP HANA tenant databases isolated, and what do they share?

How does SAP HANA persist in-memory changes, and how often by default?

What three things are always true of a SAP HANA backup?

In what order does a SAP HANA recovery restore backups?

When is a SAP HANA log segment automatically backed up?

What are the three SAP HANA backup destinations, and what is Backint?

What does multitier system replication enable beyond two systems?

What is an in-memory database (IMDB), and what hard requirement and scaling options does it impose?

Since RAM is volatile, how does SAP HANA survive a crash without losing committed data?

How does a column-store table store data, and what's HANA's main compression technique?

Why is row-based storage good for OLTP but bad for OLAP, and how does column storage flip this?

Where do SAP HANA data and log volumes reside, and which parameters define them?

How do SAP HANA processes start, and what does hdbdaemon do?

Why must you never touch SAP HANA datafiles with OS commands?

What are SAP HANA redo logs called, and why must log backups be scheduled?

What is the SAP HANA administrator OS user, and what are its groups?

What's the difference between a differential delta and an incremental delta SAP HANA backup?

Why does SAP HANA system replication preload column tables on the secondary?

What does high isolation mode do for multiple tenant databases?

How does the index server map to tenant databases, and what does the daemon do?

Why must a HANA system's HSID differ from any tenant database SID?

What is a HANA system's identifier and instance model?

What is the SYSTEMDB, what global tasks need it, and why is at least one tenant required?

How does SAP HANA system replication work, and what modes are available?

What does strict multitenancy mean in SAP HANA 2.0, and what's a key consequence?

What top directory does SAP HANA use, and what's in /usr/sap/<HSID>?

What's notable about how a SAP HANA data volume (datavolume_000.dat) grows and shrinks?

What does the HDB info command show?

What do the compile server, preprocessor server, and script server do, and which is optional?

After a crash, how does SAP HANA rebuild a consistent database?