Licensing scale up & usage of single/multiple designer nodes
Hi!
I would like to find out two things about dataiku:
1. Is there a provision of scaling up licensing for additional data scientist roles in dataiku infrastructure. Can you help elaborate on this
2. How does dataiku enable the usage of single/multiple designer nodes? What does dataiku team recommend?
Thanks
Riti
Answers
-
Manuel Alpha Tester, Dataiker Alumni, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Dataiku DSS Adv Designer, Registered Posts: 193 ✭✭✭✭✭✭✭Hi,
The software supports both of your asks, but these are commercial questions, controlled by your license:
- Go to this page and select Installed;
- Review the 4 different options (Free, Discover, Business, Enterprise);
- Click "View Edition Comparisons" for features.
I hope this helps.
-
thanks for your response Manuel. However, I'm looking for something more specific which doesn't get answered in the links you mentioned
1. If we go with multiple designer node with Dataiku
- what are the pros and cons?
- Cross Designer node access of project, data points, recipes, code
2. What are the pros and cons to handle load on huge EC2 with Single designer node?
-
Manuel Alpha Tester, Dataiker Alumni, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Dataiku DSS Adv Designer, Registered Posts: 193 ✭✭✭✭✭✭✭
Hi,
On 1), multiple design nodes typically break the collaboration. A project in node A is not accessible in node B, unless you import it explicitly;
On 2), if possible, it is preferred to keep all users in the same design instance, because you enable collaboration and reusability of all assets. If you are worried about the concurrency of many design users, there are certain limits you can configure in DSS (global, per job, project, recipe, type, connection, etc) and cgroups in Linux.
I hope this helps.
-
Greg Dataiker, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Adv Designer Posts: 6 DataikerHi Riti,
There are valid reasons why you might want to have multiple Design nodes - for example to separate projects, use-cases, teams, geographies, for compliance reasons etc. Data sources, projects, code repositories and other assets can be shared between environments. Dataiku's Cloud Stacks deployment methodology makes it simple to manage multiple instances in a consistent way.
Thanks,
Greg