-
Notifications
You must be signed in to change notification settings - Fork 0
OGDI DataLab v5 at a Glance
In order to install OGDI DataLab v5, a Windows Azure accounted with one Cloud Service (previously Hosted Service) for and two Storages (previously Storage Account) are needed. The Cloud Service is used for the DataBrowser’s web role and worker role. The two separate storages are used for:
- storing end-point information and configuration;
- storing publicly accessible data.
Make sure all three items are in the same Region/Affinity Group to optimize data transfer and speed. When creating the cloud service, specify you are not going to deploy configuration the service at that time.
The OGDI DataLab v5 project does not contain all 3rd party dependencies. You need to install the following using NuGet:
- For the DataBrowser.WebRole project, download OGDI DataLab DataBrowser.WebRole Dependencies
- For the DataService.WebRole project, download OGDI DataLab DataService.WebRole Dependencies
- For the DataLoaderGuiApp project, download OGDI DataLab DataLoaderGuiApp Dependencies
- For the DataLoader project, download OGDI DataLab DataLoader Dependencies
Take a note of the exact names as NuGet has many packages with similar names.
The DataBrowser.WebRole must contain a connection string to the configuration storage in Windows Azure. The settings in its Cloud part must contain the account credentials – Account name and Account key from the Windows Azure.
The serviceUri in the same settings must be set to the full URL given to the Cloud Service in Windows Azure, including port and extension. The format is as follows:
where myDnsPrefix is the name given to the Cloud Service when it was created.
The project works well in extra-small instance of a VM.
Same connection string elements, Uri, and VM size must be complete in the DataBrowser.WorkerRole.
To publish OGDI DataLab v5 on Windows Azure, compile DataBrowser.Cloud and publish it using Cloud service configuration. It can be done through Visual Studio or independently by uploading package files.
To use the ConfigTool, you need to define AccountName and AccountKey in the Web.Config file, linking it to the configuration storage in Azure.
Start a local instance of the ConfigTool and configure an end point in the browser. Use the AccountName and AccountKey from the storage you designated for publicly accessible data in Windows Azure.
The DataLoaderGuiApp project will allow you to upload data from CSV and KML files. Configure the connection settings / end-point first using the same account name and key for the data storage.
In the DataLoaderGuiApp, configure the dataset using the OGDI Metadata Designer form. All data in the CSV must be valid as there is no check for validity while uploading.