cancel
Showing results for 
Search instead for 
Did you mean: 

Random HANA Studio Questions

Former Member
0 Kudos

1) What's the difference between system and server? For example, in HANA studio, you can "add system", are you connecting to one server which is called a system also? Or there can be several servers in one 'system'?

2) Is it good practice to filter data at schema level or attrubite level? (when data is coming from ECC). Should the approach be to move most data and filter or filter when moving it?

3) Granularity, what does it mean? is it good or bad? I always here co-workers saying this and I'm too embarrassed to ask.

4) Do I need a universe to connect Webi to HANA Views directly? (for Crystal Report it doesn't seem to be needed)

5) If I need to create a view using a transaction table and link it with header tables, but don't need any no numeric values (measures), should I just add the transaction table to the attribute view? Or good business practice is to create all transaction tables in analytic view (regardless if it has numeric values or not/measures)?

Accepted Solutions (1)

Accepted Solutions (1)

hai_murali_here
Advisor
Advisor
0 Kudos

Hi Mic,

Your questions are very well answered by Ravi.

Just to add from my side,

1) Adding system in HANA Studio means adding one HANA Server.You can also add multiple instances of this HANA Server with different users.

2)Its advisable to filter the required data during data load into HANA,instead of moving all the data into HANA and then applying the filter on unwanted data.You have the filtering provisions in Data Services and also in SLT.

3)Granularity refers to the "details" on data.When you want to see the data in more granular level,then the data to such detailed level should exist in the system to perform further drill downs.

4)If you want to consume HANA Views in WebI,you have to go via Universe and there is no direct connectivity supported as of today.But in case of Crystal Reports they have come up with direct connectivity option in FP3 version.I guess they will provide such option in WebI too in near future.

5)Its good to keep master data in Attribute view and combine them with transactional data in Analytical View.

Rgds,

Murali

Former Member
0 Kudos

Hi Murali,

This is Srinivas, I am having 5 yrs. of Exp. in Siebel CRM. As Oracle is going to introduce Fusion to replace Siebel I am planning to change my track. I have some couple of options to think. SAP HANA is one of them. I dont have any prior experience in any of the modules of SAP. But I am considering this as its a new technology in SAP Umbrella. Can you please give your comments for the below Queries.Sorry If my Questiosn looks like meaning less and simple?

1. Having experience in Siebel learning SAP HANA is it a right decision?

2. Is it sufficient if I learn SAP HANA and SAP BO for present trends? Can I survive in realtime by having these two technologies as 2 yrs. SAP BI & SAP HANA experienced guy?

3. what is the expected scope for SAP HANA and SAP BI for down the line 10 yrs.?

4. Why cant I go for Informatica & OBIEE why should I choose SAP HANA & SAP BI in terms of market scope and job opportunities?

Thanks,

Srinivas M.

Answers (4)

Answers (4)

Former Member
0 Kudos

Thank you all, you all have been very helpful 😃

rama_shankar3
Active Contributor
0 Kudos

Mic:

On your webi/universe question: yes, you will need a universe to connect to HANA tables. It is not a limitaton of HANA but a requirement for Webi.

Regards,

Rama

Former Member
0 Kudos

Thank you Ravindra and Murali. I have follow up questions:

Sorry, I'm still new:

1) What's an instance? What's the difference between a server and an instance?

2) I'm still little unsure about using how to use a transactional table when I don't need any values/measures from it. For example, Analytic view forces me to select a measure from a table, but if I'm not using any measures what should I do? (it won't let me save/activate)

rama_shankar3
Active Contributor
0 Kudos

Mic:

Below are my thoughts on your questions:

1 -   One server can have multiple instances in HANA. For example 1  SAP HANA server can host 1 or more database instances. Usually, it is one-to-one but depending on customer needs we can have multiple instances of databases in one server.

2 - You can assign a dummy constant with value 0 or 1 if you do not have any measures in the data set

Hope this helps.

Regards,

Rama

former_member184768
Active Contributor
0 Kudos

Hi Mic,

From any database perspective, it consists of "Software" and "memory" (physical files / RAM). The Software manages the memory which contains data and the software provides access to read / write the data to memory.

Instance is nothing but "instantiation" of the software to operate on the memory. You can have multiple instantiations of the software to operate on multiple memory segments independent of each other.

Each instance represent a Memory area (Physical / RAM) and a port to which the services are listened to and identified by the instance id.

Thus, one physical server can hold multiple instances provided they operate on independent memory areas and have different ID / PORTs.

Coming to the second question about Transaction data without measure, it is called as "Factless Fact" in Data warehousing terminology. It is a practical scenario and to implement it in HANA, you may have to use a dummy fact (constant value). There could be alternative of using it as Master data, but it depends on the business implementation and nature of the data.

Regards,

Ravi

Former Member
0 Kudos

Hi Rama,


This is Srinivas, I am having 5 yrs. of Exp. in Siebel CRM. As Oracle is going to introduce Fusion to replace Siebel I am planning to change my track. I have some couple of options to think. SAP HANA is one of them. I dont have any prior experience in any of the modules of SAP. But I am considering this as its a new technology in SAP Umbrella. Can you please give your comments for the below Queries.Sorry If my Questions looks like meaning less and simple?


1. Having experience in Siebel learning SAP HANA is it a right decision?

2. Is it sufficient if I learn SAP HANA and SAP BO for present trends? Can I survive in realtime by having these two technologies as 2 yrs. SAP BI & SAP HANA experienced guy?

3. what is the expected scope for SAP HANA and SAP BI for down the line 10 yrs.?

4. Why cant I go for Informatica & OBIEE why should I choose SAP HANA & SAP BI in terms of market scope and job opportunities?



Thanks,

Srinivas M.

Former Member
0 Kudos

Hi Ravindra,


This is Srinivas, I am having 5 yrs. of Exp. in Siebel CRM. As Oracle is going to introduce Fusion to replace Siebel I am planning to change my track. I have some couple of options to think. SAP HANA is one of them. I dont have any prior experience in any of the modules of SAP. But I am considering this as its a new technology in SAP Umbrella. Can you please give your comments for the below Queries.Sorry If my Questions looks like meaning less and simple?


1. Having experience in Siebel learning SAP HANA is it a right decision?

2. Is it sufficient if I learn SAP HANA and SAP BO for present trends? Can I survive in realtime by having these two technologies as 2 yrs. SAP BI & SAP HANA experienced guy?

3. what is the expected scope for SAP HANA and SAP BI for down the line 10 yrs.?

4. Why cant I go for Informatica & OBIEE why should I choose SAP HANA & SAP BI in terms of market scope and job opportunities?



Thanks,

Srinivas M.

former_member184768
Active Contributor
0 Kudos

Hi Mic,

I'll try to reply to the questions to best of my understanding:

1) In HANA studio, each system represent a server. The connection parameters in adding the system point to a server URL. It may have multiple blades, but I think one server only.

2) More data in the system adds to the server sizing in HANA, which means more cost. But if you see the need for the data in future, then you can always load it in the system. So if the initial need is to load the required data, then it should be filtered before moving into HANA. Provided the subsequent data load does not impact the initially loaded data.

3) Granularity is the level of data detail. Higher granularity means more detailed data. It is good to have granular data when the reporting is required at the detailed level. But if the reporting is required at the summarized level, then it is not required to load the data at higher granularity. As I mentioned earlier, more data means more memory, which is higher cost.

4) My knowledge on BO tools is limited, but I think you need the universe to connect to HANA views in WebI. You can build the universe even on the tables in HANA. But it is advisable to build it on the views, as the business logic can be implemented on the views pushed down to the database rather than implementing it in the universe.

5) If the transactional data is being used as MASTER data, then it can be used for Attribute view. The Analytic view represent the InfoCube in BW, hence is should have Numeric values as measures.

Hope this helps.

Regards,

Ravi