HANA SP9 – First impressions
Hello HANA community,
Now that SP9 has been released, I figured I would finally write my first blog post with a few comments about this new release.
There is a post here with a nice summary of what’s new in SP9, in case you haven’t seen it yet.
I run a VM on my home computer with SLES 11 SP3 for HANA testing. In case you are wondering about the specs, it’s an i7 with 32GB ram plus a few SSDs. The performance is good enough to use as a sandbox/playground.
The first thing I noticed is that the SP9 download size is ~43% smaller than the SP8 revisions:
As far as I can tell, the biggest size reduction happened with the server component itself, that went from 4,72 GB to 2,18 GB. I’m sure there’s a perfectly reasonable explanation for this. 🙂
Some packages were removed and some were added, here’s a comparison with the changes:
Strange that the River package was removed. Maybe it was integrated on some other delivery unit?
EDIT: River was broken down and parts of it were integrated into the HANA Platform. More details here.
According to note 2001528, starting with revision 80 you should have GCC 4.7 installed. SP9 enforces this with a check when you start the installation. The instructions for the manual update are on the note.
You can now select the database mode during setup – single or multi container (for multiple tenant databases):
The setup did not warn that I was running an unsupported configuration. I noticed this also happened on SP8. Maybe the hardware check script now gives you a “free pass” when you are running VMWare because virtualization is officially supported for productive environments?
Another issue that started happening on SP8 was that setup would fail when importing delivery units. This has now been fixed, and everything is correctly imported.
Total install time was 21 minutes. I don’t have a reference duration for SP8, but I have the feeling SP9 was significantly slower.
After installing the database, I manually installed HANA Studio. The version that ships with the SP9 install is 2.0.7:
The new HANA Studio setup now lets you select what components to install. Answers is already included, cool:
Now when you add a system, you need to specify the database mode and tenant name, if necessary:
A new (I think?) Fiori app called SAP HANA Cockpit is available from the context menu inside HANA Studio:
It’s a nice Fiori app that gives us a system overview.
The Web IDE is looking really good! Continuing to grow into a full-fledged web version of HANA Studio:
The HANA Lifecycle management web app also has a few enhancements:
One of the new delivery units present on the SP9 install – SAP_DB_CONTROL_CENTER – contains this app:
It appears to be a cockpit to manage multiple HANA systems (maybe multiple tenants?)
I also tried to run the spatial demo but the NOKIA HERE maps developer website wasn’t in the mood for collaborating. 🙂
This was a just a brief run through a fresh SP9 install. I’ll be investigating the new features in detail over the next few days – in particular the smart data streaming and dynamic tiering, see what it’s all about.
Overall, this new release seems to be a big step in the right direction.
For now, my favorite new features are:
– Multi-tenant database containers;
– The Hadoop controller!
What are your favorite new features in HANA SP9?