gocosmos - REST client

gocosmos driver uses its own REST client to communicate with Azure Cosmos DB SQL API. The REST client can be used as a standalone package.

The REST client supports:

• Database: Create, Get, Delete, List commands and changing throughput.
• Collection: Create, Replace, Get, Delete, List commands and changing throughput.
• Document: Create, Replace, Get, Delete, Query and List commands.

Example usage:

package main

import (
	"github.com/microsoft/gocosmos"
)

func main() {
	cosmosDbConnStr := "AccountEndpoint=https://localhost:8081/;AccountKey=<cosmosdb-account-key>"
	client, err := gocosmos.NewRestClient(nil, cosmosDbConnStr)
	if err != nil {
		panic(err)
	}

	dbSpec := gocosmos.DatabaseSpec{Id: "mydb", Ru: 400}
	result := client.CreateDatabase(dbSpec)
	if result.Error() != nil {
		panic(result.Error())
	}

	// database "mydb" has been created successfully
}

Connection string syntax for Cosmos DB

Note: line-break is for readability only!

AccountEndpoint=<cosmosdb-endpoint>
;AccountKey=<cosmosdb-account-key>
[;TimeoutMs=<timeout-in-ms>]
[;Version=<cosmosdb-api-version>]
[;AutoId=<true/false>]
[;InsecureSkipVerify=<true/false>`]

• AccountEndpoint: (required) endpoint to access Cosmos DB. For example, the endpoint for Azure Cosmos DB Emulator running on local is https://localhost:8081/.
• AccountKey: (required) account key to authenticate.
• TimeoutMs: (optional) operation timeout in milliseconds. Default value is 10 seconds if not specified.
• Version: (optional) version of Cosmos DB to use. Default value is 2020-07-15 if not specified. See: https://learn.microsoft.com/rest/api/cosmos-db/#supported-rest-api-versions.
• AutoId: (optional) see auto id section.
• InsecureSkipVerify: (optional) if true, disable CA verification for https endpoint (useful to run against test/dev env with local/docker Cosmos DB emulator).

Known issues

GROUP BY combined with ORDER BY is not supported

Azure Cosmos DB does not support GROUP BY combined with ORDER BY yet. You will receive the following error message:

'ORDER BY' is not supported in presence of GROUP BY.

Cross-partition queries

When documents are spanned across partitions, they must be fetched from multiple PkRangeIds and then merged to build the final result. Due to this behaviour, some cross-partition queries might not work as expected:

• Paging cross-partition OFFSET...LIMIT queries using max-count-item:
  Since documents must be fetched from multiple PkRangeId, the result is nondeterministic. Moreover, calls to RestClient.QueryDocumentsCrossPartition(...) and RestClient.QueryDocuments(...) without pagination (i.e. set MaxCountItem=0) may yield different results.

• Paging cross-partition ORDER BY queries with max-count-item:
  Due to the fact that documents must be fetched from multiple PkRangeId, rows returned from calls to RestClient.QueryDocuments(...) might not be in the expected order.
  Workaround: if you can afford the memory, use RestClient.QueryDocumentsCrossPartition(...) or RestClient.QueryDocuments(...) without pagination (i.e. set MaxCountItem=0).

• Paging SELECT DISTINCT queries with max-count-item:
  Due to the fact that documents must be fetched from multiple PkRangeId, rows returned from calls to RestClient.QueryDocuments(...) might be duplicated.
  Workaround: if you can afford the memory, use RestClient.QueryDocumentsCrossPartition(...) or RestClient.QueryDocuments(...) without pagination (i.e. set MaxCountItem=0).

• GROUP BY combined with max-count-item:
  Due to the fact that documents must be fetched from multiple PkRangeId, the result returned from calls to RestClient.QueryDocuments(...) might not be as espected.
  Workaround: if you can afford the memory, use RestClient.QueryDocumentsCrossPartition(...) or RestClient.QueryDocuments(...) without pagination (i.e. set MaxCountItem=0).