atlas-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Radley <david_rad...@uk.ibm.com>
Subject Meeting on 25th of April about relationships design
Date Tue, 02 May 2017 17:26:42 GMT
Hello,
I thought I would share details of some very productive meetings I 
attended about Atlas relationships.

On the 25th of April,
We discussed the Jira 1690 
https://issues.apache.org/jira/browse/ATLAS-1690 on relationships. I think 
this Jira is an enabler for a lot of capability including a new glossary 
https://issues.apache.org/jira/browse/ATLAS-1410.
  The meeting attendees included committers and a PMC member (Madhan)

anaik@hortonworks.com, amestry@hortonworks.com, David 
Radley/UK/IBM@IBMGB,, Graham Wallis/UK/IBM@IBMGB,, Mandy 
Chessell/UK/IBM@IBMGB, Nigel L Jones/UK/IBM@IBMGB, 
ssubramanian@hortonworks.com, svenkat@hortonworks.com, 
sshivaprasad@hortonworks.com. "Madhan Neethiraj" <madhan@apache.org>, 


I walked through the relationships design. We had a vibrant discussion 
around the approach to take. We are thinking that in addition to a 
technical document, it would be useful to story board how this would 
effect users, taking into a account user interfaces. We are looking to 
pursue both these approaches. On the technical front, I think more 
examples of how these relationships would look in metadata , instance 
level and in the graph would be helpful.

For these story board - higher level stories, there are projects detailing 
proposals how metadata can/should be used by many different governance 
roles. These projects are proposals for inclusion in Atlas. Various 
companies have contributed to the thinking around these projects; I would 
like to explicitly invite the community to feedback/ contribute to these 
projects. The projects are being documented in the newly opened up Atlas 
wiki :  https://cwiki.apache.org/confluence/display/ATLAS/Atlas+Projects 

On the 2nd of May
  The meeting attendees included committers and  PMC members (Suma and 
Madhan)

anaik@hortonworks.com, amestry@hortonworks.com, David 
Radley/UK/IBM@IBMGB,, ,  ssubramanian@hortonworks.com, 
svenkat@hortonworks.com, sshivaprasad@hortonworks.com. "Madhan Neethiraj" 
<madhan@apache.org>, eostic@us.ibm.com

We reviewed the v1.5 design document for Jira 1690. Madhan suggested the 
attribute names of the end points were held in the relationship object and 
not in the the typedefs, we all felt this simplified the model. We agreed 
not use structures in the relationship as they require a separate vertex; 
instead we would allow simple attributes. We talked of the relationship 
being a vertex to cut down the number of edges especially for many to many 
relationships. We agreed to add cardinality to the end points. We agreed 
to remove iscontainer from the relationshipdef endpoint and have a 
containerendpoint in the relationshipDef. We agreed there was enough 
consensus to start coding. We agreed that we would look at the search use 
cases then move onto the glossary. We talked about Ranger integration 
being important.

I look forward to hearing your thoughts,ideas and responses,     all the 
best, David. 

 
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU


Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message