TM1 Forum

Hi

If someone has already done this before, I would appreciate a side-by-side comparison of Cubeware's importer vs TM1's TI with respect to

1. read/writing into TM1 cubes,
2. changing TM1 dimensions,
3. changing TM1 security,
4. allowing drill thru from TM1 cells,
5. speed of loading/changing,
6. being able to access multiple datasources at the same time,
7. Error trapping

Many thanks.

Here's the hoping

Cheers

Kaveenga

Hi Kaveenga,

1. Cubeware Importer can read from and write to TM1 cubes - as it can do from/to many other OLAP databases like PALO, Infor PM, MSAS, Essbase, PowerOlap and more. Importer can also read from and write to relational sources.
2. Importer is capable of creating/updating dimensions in all these OLAP databases (incl. TM1).
3. You can use Importer to set security (in fact loading the security cubes).
4. Importer is a data integration tool. Drill-through would be initiated from a reporting/analysis frontend - for Cubeware this would be Cockpit, which is capable of doing a drill-through to source data. Drilling through from the TM1 Server Explorer cube viewer to the source data using Cubeware Importer instead of TI is not possible.
5. Cubeware Importer has a reasonable upload speed, but it does not come near the TI upload speed.
6. With Importer you can access multiple datasources - a single 'mapping' (the Importer 'process') can process f.i. a text file and simultaniously also do lookups on relational tables and (TM1 and non-TM1) cubes and dimensions.
7. An Importer mapping creates a log-file stating warnings and fatal errors. Most mapping objects have error handling settings. You can also f.i. use filter objects, logging object, etc. to set up custom error handling.

I've done quite some projects with Cubeware Importer and Cockpit, most on top of PALO, but also on TM1. Overall the strenghts (opposed to TM1) from my experience for Importer are the 'visual' interface which is experienced by our customers as easy to understand and use by (business) end users.
You can process data from server to server, where the one server could be a TM1 server and the other a non-TM1 server. Importer has much better scheduling possibilities than TI.
Weaknesses opposed to TI are the update speed, which is considerably slower than TI and the fact that it is not an fully integrated part of the solution (see f.i. point 5. above).
Regarding speed I can add that in the business cases that we used Importer the ease of use and the single interface to multiple sources was outweighing the upload speed, which is 'good enough' in most cases. And if you're doing massive data uploads you can still start a TI process from an Importer job...

HTH, Michel

Edit: after posting this message I found that it was actually in the PALO forum...

Hi Michel,

Many thanks for this post. The ability to start a TI process from Importer is an interesting fact.

Kind regards,

Kaveenga