Range Index pre-existing collection programmatically - azure

Question

I've created a database with a collection. The collection has thousands of pre-existing documents which looks something like below as an example.
{
"Town": "Hull",
"Easting": 364208,
"Northing": 176288,
"Longitude": -2.5168477762,
"Latitude": 51.4844052488,
}
I'm aware that I need to index the database with a range type so I can use the range query & the OrderBy function with my data.
So, how can I range index the pre-existing data programmatically using the .NET SDK?
I've come up with the below code. However, it seems to fail at querying the collection. When I've inserted a breakpoint the 'database' contains null at the point of querying for the collection.
// Create an instance of the DocumentClient.
using (dbClient = new DocumentClient(new Uri(Properties.Settings.Default.EndpointUrl), Properties.Settings.Default.AuthorizationKey))
{
Database database = dbClient.CreateDatabaseQuery().Where
(db => db.Id == Properties.Settings.Default.databaseID).AsEnumerable().FirstOrDefault();
DocumentCollection collection = dbClient.CreateDocumentCollectionQuery(database.SelfLink).Where
(c => c.Id == Properties.Settings.Default.collectionID).ToArray().FirstOrDefault();
// If database type is not null then continue to range index the collection
if (collection != null)
{
stopsCollection.IndexingPolicy.IncludedPaths.Add(
new IncludedPath
{
Path = "/*",
Indexes = new System.Collections.ObjectModel.Collection<Index>
{
new RangeIndex(DataType.String) {Precision = 6},
new RangeIndex(DataType.Number) {Precision = 6}
}
}
);
}
else
{
Console.WriteLine(">> Unable to retrieve requested collection.");
}
}

Answer 1

Today, indexing policies are immutable; so you will need to re-create a collection to change the index policy (e.g. add a range index).
If you wanted create a collection with a custom index policy programatically, the code to do this would look something like this:
var rangeDefault = new DocumentCollection { Id = "rangeCollection" };
rangeDefault.IndexingPolicy.IncludedPaths.Add(
new IncludedPath {
Path = "/*",
Indexes = new Collection<Index> {
new RangeIndex(DataType.String) { Precision = -1 },
new RangeIndex(DataType.Number) { Precision = -1 }
}
});
await client.CreateDocumentCollectionAsync(database.SelfLink, rangeDefault);
And then write some code to reads data from the existing collection and writes the data over to your new collection.
But this is a bit cumbersome...
As an alternative solution... I would highly suggest using the DocumentDB Data Migration Tool to create a new collection with your new index policy and move data from your old collection to the new collection. You can delete the old collection once the migration completes successfully.
You can download the data migration tool here.
Step 1: Define DocumentDB as source:
Step 2: Define DocumentDB as the target, and use a new indexing policy:
Hint: you can right click in the indexing policy input box to choose an indexing policy
which will give you an indexing policy that looks something like this:
{
"indexingMode": "consistent",
"automatic": true,
"includedPaths": [
{
"path": "/*",
"indexes": [
{
"kind": "Range",
"dataType": "Number",
"precision": -1
},
{
"kind": "Range",
"dataType": "String",
"precision": -1
}
]
},
{
"path": "/_ts/?",
"indexes": [
{
"kind": "Range",
"dataType": "Number",
"precision": -1
}
]
}
],
"excludedPaths": []
}
Step 3: Run the import job...
Reminder: Delete the old collection after the import finishes successfully.