These days, you’re not done once you have selected a new EPM solution. You also have to select a tool for your database-management system (DBMS). Take CCH Tagetik, for instance. You can run it on various DBMSs such as SAP HANA, MS SQL, Oracle, and Postgres. Of these, SQL is the most widely used, but HANA could well be the DBMS of the future. Gabor Wieringa delves into both options for us. What are the consequences for performance? And how do you make the right choice based on your situation?
SAP HANA is still coming on the scene, so to speak. It has been around for some time, but it is now slowly entering its mature phase. The expectation is that this new DBMS will have an impact on the entire IT organizational landscape.
But what is the actual difference between SAP HANA and SQL, exactly? This has to do mainly with how each one reads data; in-memory versus disk. And there is also a major difference between them in terms of data storage. SQL uses row-based storage: data is arranged horizontally. SAP HANA, by contrast, uses columnar storage, which can help you access your data faster. In the right circumstances, SAP HANA can really improve performance.
It is thus important for the applications to have been developed for columnar use. If the software is still written on the basis of SQL statements, you are not using the power of HANA to the full.
But what are the aforementioned “right circumstances”? And how do things stand at the moment with CCH Tagetik? To find out, let’s head over to our very own Swap Support Test Center, where we put new applications through their paces, in both functional and technical terms, before we install them in an organization. That way, we know what the consequences of a given choice are on stability and performance, for instance.
Within our own SAP HANA test environment, where we have been testing SAP applications for some time, we have also looked closely at CCH Tagetik. We did this by running the same consolidation on both SAP HANA and SQL. And we always share our findings with CCH Tagetik, so that their developers can use our test results for the next release.
What is striking is that a clear progression can be observed. About 18 months ago, SAP HANA was still having a lot of teething troubles, which happens quite a lot with the first few releases of a new technology. In our first tests, it was even difficult to complete the consolidation and the reports associated with it.
However, we are seeing that, with each release, performance improves by about 30%. At the moment, the SAP HANA database in the tests is not yet faster than the SQL database. We do expect this to change, though: development is happening so fast that we expect SAP HANA to perform much better soon.
It is worth mentioning, too that, with some tweaking and fine-tuning, the performance of HANA can be improved even now. However, this also applies to SQL; here, too, fine-tuning makes for improvements in performance.
At the moment, SAP HANA’s strength lies mainly in how it processes data. It is therefore entirely possible that the improvement in speed happens more when the data is being transferred from the source system to CCH Tagetik. HANA can also be useful if you want to use multiple source systems. The database-management layer can then load all the data from the individual systems into CCH Tagetik.
Also coming on the scene is the Analytical Information Hub (AIH), which allows you to analyze big data and run what-if scenarios. We see a nice role here for SAP HANA: you perform these heavy calculations in-memory, and the power of HANA comes into its own
Ultimately, the choice of a DBMS is about performance and how hefty the calculations needed are. If the focus is on consolidation and reporting, SAP HANA may be a bit premature at the moment. It would be a better idea in that case to wait until developments have come along a bit farther. If the focus is on the number of records and analyses – for instance because you are processing a lot of transactional data – then SAP HANA is a great option, even now.
Well, as you can see, the choice is not that easy. A DBMS must fit in well with your applications, your overall landscape, and your organization. It helps to pre-test carefully, so you know what impact to expect with which choice. This is especially important where a solution is still under development and each new release yields different results, so you can make an informed choice.
Our Test Center is available for just that. Gabor will also be delighted to answer any particular questions you have. Just drop Gabor a line.
Would you like to keep abreast of developments related to SAP HANA and CCH Tagetik? Just sign up here and we will be happy to send you the latest test results.
Text: Gabor Wieringa