A lot of customers asked me in the last weeks 'what should be used HANA 1.0 or HANA 2.0?'. There are some reasons you can't avoid to migrate to HANA 2.0, but there are still some aspects left resulting in staying at HANA 1.0.
There are tons of new features and improvements with HANA 2.0, but do you
really need them?
The most common answer against HANA 2.0 is the complexity coming with
MDC. For sure you have to review your concepts for running HANA systems regarding security/authorizations, backup&restore, parametrization and monitoring. SAP will improve such tasks and make the switch easier, but it is currently a challenge to adapt it.
When
should and when I
have to use HANA 2 or even it is
not wise to upgrade?
The 'should' section is more like
nice to have but may not really important for your business.
The 'have to' section includes aspects which you can't
avoid or influence if they come true.
The 'should not' section listing all characteristics
pro HANA 1.0.
For sure the list is not complete and there are more aspects and I will adjust it if there are more important ones, but it only should give you a first overview and tendency.
You
should use it when you...:
- want to benefit from new performance features like Dynamic result Cache, improved garbage collection / locking, optimized compression, HEX (HANA Execution Engine), ESX (Extended SQL Executor)
- want to use event triggered log backups
- want to install license keys for individual tenant databases
- want to use Data Shipping Parallelization within HSR
- want to use redo log encryption (1642148)
- want to restore a single database backup directly into a tenant database (
- want to lower your TCO with persistent memory (pmem) / NVM / NVRAM (2618154)
- want to use HANA Cockpit for monitoring and managing your HANA databases (some features are only supported in connection with HANA 2.0)
You
have to use it when...:
- it is planned to use S/4 HANA 1709+ (2524661)
- it is planned to use BW/4 HANA 2.0 (2347382)
- it is planned to use HSR active / active (2391079)
- it is planned to use POWERPC SLES12 LE (little endianness)
- you want to use special functions/features in SDA / SDI / SDQ (more supported remote databases and functions/syntax)
- you want to migrate from INTEL to PowerPC or vice versa via restore/recovery (1642148)
- you want use Cross-SID-copy: Backint-based backups of a tenant database ca be recovered into a tenant database using a target SID or tenant NAME that is different to the original one (2096000)
- you want to use HANA Cockpit 2.0 (running only on HANA 2.0)
- it is planned to use secondary time travel feature with HSR (HANA 2.0 SPS03)
- it is planned to use multi-target replication (HANA 2.0 SPS03)
You
should not use it when...:
- you are currently on old hardware: Haswell and older => recommended to buy new HW (2399995)
- you are using SLES11 SP4 BE (big endianness) with PowerPC and don't want to migrate (till end of maintenance of SLES11 SP4)
- currently no feature used and not planned to use which only comes with HANA 2.0
- you want to avoid using MDC (default mode since HANA 2.0 SP01)
- you are using XSC (XS Classic) and not ready to use XSA (XS Advanced) - XSC is still supported but not recommended (2396214)
- you are using a deprecated feature (2425002)
- your current release / product is not supported on HANA 2.0 (see PAM)
- you are affected by the limitation using cross tenant access (2196359 / 2312583)
Another question which also comes up is regarding
stability. This depends on how do you use your database? Which features are used and how proper you followed the best practices (system setup / parameters) of your hardware and software vendors. Do you frequently review the
known issues of your SPS? May be there are some workarounds to get your revision stable.
For the major issues you can check the SAP HotNews collection for every SPS:
For details what will be changed in each SPS check
help.sap.com and the
HANA Academy channel