Thank you very much for your post. We are currently trying out Wherescape. My assessment of the tool is similar.

What I find negative is that when you switch to an other database system, you have to re-map all the ETL process in Wherescape. Also, there is no official API for importing and exporting the proprietary metadata.

What is positive is that you can extend the WS metadata system with custom properties for objects and columns. This allowed us to build our own templates for multiactive, timetracking, etc.

Drag & drop plus data lineage for (semi-)automatic porting of an existing data warehouse to Wherescape is very time consuming, because for Wherescape object names have to be unique across multiple schemas. I use my own metamodel in XML for porting and generate with XSLT the code for WS Red as well as updates on the WS metadata.

Many greetings, Bert

one hell of an interesting blog entry, thank you very much! And it relates to a problem zone I quite often struggle with: How to handle business keys refering to a business object from many different data sources each having its own business key range (that might overlap to the business key ranges of other data sources).

An appropriate solution for hub lemma 2 is difficult to develop. Even if I understand your statement „If I add the source system name as a prevention measure to everything, the whole concept of data integrating with data vault leads into absurdity and makes things over-complicated.“ I would take up the cudgels for option 2 if you have different data sources with overlapping business key ranges (as option 4 would require additional DDL effort to implement new data sources in a Data Vault model which in my opinion would break the advantages of Data Vault far worse than option 2). Combined with option 5 we could still align business keys from different systems by a same-as link or role-playing links…

Therefore, I plead for mercy for Hub lemma 2 option 2 🙂

The delta store is by itself not a bad thing and performs okay. It will automatically transform into a column store when it hits over 1 million rows.
But I do maintenance and transform everything into column store. In our case in a less demanding time slot at lunch time. Could do also in the night. ALTER INDEX [index_name] ON [table_name] REORGANIZE WITH (COMPRESS_ALL_ROW_GROUPS = ON). To find your indexes loop through SELECT * FROM sys.indexes WHERE type = 5.

Haben Sie das schon gelesen,
http://www.srf.ch/radio-srf-1/auf-den-spuren-eures-namens/zemp

MfG
markus Christ

Druck mal auf beide Papiere und vergleiche sie. Man sieht doch deutlich, dass das billige Papier nicht gut aussieht und auch nicht so weiss ist. Und den Eindruck als billiger Mensch daherzukommen möchtest du bestimmt nicht hinterlassen, oder?

Besten Dank!

Leider kann MacStammbaum das nicht. Ich hab bei den Entwicklern nachgefragt und es hieß, dass es nicht in der Roadmap steht da kein offizielles Feld. Leider kann man auch keine benutzerdefinierten Felder einbauen, wie es andere können.

Reunion für den Mac kann das.

Die Vasa in Stockholm hatte mich auch beeindruckt, als ich vor einigen Jahren dort war. Auffällig, wie hoch der Schwerpunkt ist (vom Sehen her!!!) Es erweckt den Eindruck, dass, Kanonen geladen, Kentern programmiert sei! Muss eine herbe Enttäuschung gewesen sein, die wunderschöne „Handarbeit“ so zu verlieren!

gruß aus deutschland