An implementor view on Pentaho tools.
February 23, 2016
An implementor view on Pentaho tools.
Score 7 out of 10
Vetted Review
Verified User
Overall Satisfaction with Pentaho
We used the Pentaho Data Integration tool to pre-process some data and fill thr production database. It was used in the department where I worked. Being a software developer I made a decision to use PDI instead of writing scripts on Python. The main advantage of PDI was supporting scaling out of the box, and the ability to run on multiple machines. When you need something less than a Hadoop cluster but bigger than self-made quick scripts it fits well.
- The main strength of Pentaho (PDI) is that it is free. No need for additional license approvals to implement something quickly using the UI tool.
- It has Pentaho Reporting - the ability to use reporting on top of PDI transformations is a powerful feature. PDI can handle data extraction, and with reporting you can create a simple report on this data. They all just work with each other without additional integration required.
- And most of Pentaho projects are open-source. That helps a lot when debugging something.
- Help documentation is an area for improvements. A lot of features and side effects just not documented.
- PDI uses swt. It has a lot of troubles with this dependency. It is run on Java, but with swt - and you are stuck with a platform that supports swt. ARM architecture does not have swt support out of the box.
- Online jobs/transformations creation. Jobs/transformations are just xml. So the lack of online editor is just 'it was not implemented yet'.
- Strange /Pentaho-solutions placement. Application is not deployed on a server in a standard way. Instead - a server should be embedded into a 'Pentaho solution' - very strange architecture.
Did not have any other products similar to what Pentaho offers out-of-the box for free. The closest was to write some scripts manually so in our case PDI has beat Python scripts.