Our Introduction to Centerprise webinar, which aired on November 6, was gratifyingly well attended. Thank you to all our customers and prospective customers who joined us to learn about the new features and functionalities in Centerprise. Many great questions were generated during the webinar, and we thought we’d share them with our blog audience. If you missed the webinar, you can view the video through our Astera TV portal, either on YouTube or Vimeo. We’ll be happy to answer any more questions you may have that did not get logged in the webinar or that come up as you view the video–just send us an email at firstname.lastname@example.org.
And don’t forget we’re running a series of in-depth webinars on best practices for Centerprise 6, starting with Working With the High Volume Data Warehouse in Centerprise this Wednesday, December 11 at 8:00 PST, 11:00 EST. Register here>>
Here are the Q and As from the Introduction to Centerprise 6 webinar:
Q1 : When it comes to data processing, at what point does data volume affect Centerprise’s performance? 100,000 , 1 million, 10 million records?
A : In general, Centerprise’s performance is linear. However, sorting large datasets may cause Centerprise to slow down. Minimize data sorting whenever you can. (large = > 10 million)
Q2 : Are there any changes that we need to make or that you recommend we make to our V5 workflows and/or dataflows with regard to parameter passing or shared expressions or shared connections?
A : Version 6 is backwards compatible. You do not need to make any changes. All best practices involving shared items are still in place for version 6. The only thing you might want to take a look at is the ability to set variables from within a dataflow. If you have used a temporary file or database to store values across dataflows, you may want to look into variables.
Q3 : I saw your blog about name and address parsing. Can you touch briefly on that?
A : Centerprise’s name and address parsing functionality comes with the USPS database, which is updated quarterly. It is used to parse, correct, and enrich name and address data. Name parsing is used to parse names into first name, middle name, last name, suffix, prefix, etc. Address parsing and correction is used to standardize incorrect or missing data from postal addresses.
Q4 : For the processing of data flows, is there a finer granularity in the job log to see where in the dataflow it is spending its time?
A: We have added more granularity in our progress window. Now it shows time spent per record for each of the actions in the flow.
Q5: Where would I use Source as Transformation?
A: Source as Transformation is useful when you want to use a map to supply a source parameter. For example, for a delimited source file, you can supply the file path using a map. In this case, a file path input will result in an output batch of source records.
Q6: Can I compare values in my current record with those in the previous record?
A: Yes. You can use expression variables to achieve this.
Q7: Do you have a full list of new features?
A: Yes. You can download the datasheet from here – http://www.astera.com/centerprise/datasheet
Q8 – Are there any new connectors to source systems like JDE, Lawson, Oracle or SAP?
A: Not at this time.
Q9 – Is the master data capability general or does it have source system identification?
A: The user is required to build logic for de-duplication.
Q10 – Is V6 is fully compatible with V5 data flow?
Q11 – What are the benchmarks on the changed dataflow?
A: We’ll be publishing version 6.0 benchmarks through a blog soon.
Q12 – What are the resource costs for upgrades?
A: Since version 6.0 is completely backward compatible, there is negligible cost associated with the upgrade.
Q13- Is there an “end to end” user guide document?
A: Documentation for the additional features in 6.0 is in progress. It should be available by early January.
Q14 – Universal flag on restart job – Yes or no?
Q15 – How many cores are used by each dataflow?
A: It depends on the structure of the dataflow. If the dataflow in question is conducive to multiple threads running in parallel, the flow can use all the cores.
Q16 – Is the job log stored in backend tables?
A: Job log is stored in a database table.