Jerome Pineau
posted this on October 29, 2009 06:51
This FAQ is designed as a repository of the most common questions we receive from both ISV (OEM) and SI (integration partners) and is moderately technical. It is a work in progress and will be updated accordingly over time.
1. Can I embed both dashboard and reports?
2. Do you handle SSO and if so how?
Our platform authentication is token-based using cookies with periodic expiration mechanisms. If your system uses token based authentication or a standard like LDAP, OpenID or OAuth, we can integrate with your security management layer and provide seamless SSO authentication for your users. Basically we treat your authentication layer as a provider into our system and the experience is seamless for the user.
3. How does GoodData handle security?
On the wire everything is SSL over HTTP. Sign-on and authentication use SAML 2.0. At the logical level, the security perimeter is project level, so users in a project can never see into other projects (and each project has database instance affinity). From an internal infrastructure standpoint, GoodData follows redundant enterprise-level multi-layered control, audit and protection protocols. GoodData is also Salesforce Certified and actively working on SAS 70 I certification. The platform is hosted on Amazon EC2 which is PCI level 2, 3, and 4 compliant and SAS 70 II certified.
4. What happens if my upload encounters a problem?
If you have selected to received email notifications of upload completion, an error message will be included in the email you received. If not, you will see an error indicator in your Data section's status column. In this example, all uploads were successful.
4. Can I physically delete my data from your platform?
When you delete a project, all data within the project is logically deleted. It is subsequently garbage-collected and physically wiped from storage.
5. How can we administer client roles and access levels?
All clients are isolated at the project level. Once invited into a project, a client's access level (role) can be set to Editor or Admin, or disabled altogether. This granularity of control (ie: role management) is available via the API. In the application and at the report level, GoodData implements "variables" which control report viewing on a per-user basis. The variables can work either as numerical or filter entities (so for example, a report on project peers may show or hide specific users from one another by filtering certain groups out, or display report metrics based on user-specific variable). Access to setting these variables is also possible via the API.
A variable is defined to take on a different value depending on the user inside this project. In this example, only one user can see report information pertaining to Abigail, Abraham and Adam. All user management functionality available via the UI is also available programmatically via the API.
6. How fast does it take to generate reports?
Performance typically depends on the size of your fact table(s). As a very rough guideline, our own FoodMart demo project which contains 2.4 million rows of sales facts typically scans in about 5 seconds. The best way to properly determine performance in any given situation is to load and benchmark it with real data.
7. Can we write our own front-end?
Definitely. GoodData's UI leverages our API exclusively for all its features and functionality.Can we control the look and feel of the GoodData UI?
You can control many aspects of report and dashboard look and feel from colors, label orientation, to size and positioning. At the moment, there is no generic way to control the skinning or other look and feel of the application (including positioning of controls and widgets). Flexible branding options are available (for additional information, please contact info@gooddata.com)
8. Can we benchmark and audit usage across multiple clients?
It's possible, and we actually do this internally for administrative and reporting purposes. You would need to use the API to pull relevant data from each client project and aggregate it into the necessary statistics for benchmarking or logging/auditing.