GridGain Developers Hub

.NET Client

GridGain 9 clients connect to the cluster via a standard socket connection. Unlike GridGain 2.x, there is no separate Thin and Thick clients in GridGain 9. All clients are 'thin'.

Clients do not become a part of the cluster topology, never hold any data, and are not used as a destination for compute calculations.

Getting Started

Prerequisites

To use C# thin client, .NET 6.0 or newer is required.

Installation

C# client is available via NuGet. To add it, use the add package command:

dotnet add package GridGain.Ignite --version 9.0.0-g4d770d8f1f

Connecting to Cluster

To initialize a client, use the IgniteClient class, and provide it with the configuration:

var clientCfg = new IgniteClientConfiguration
{
  Endpoints = { "127.0.0.1" }
};
using var client = await IgniteClient.StartAsync(clientCfg);

Authentication

To pass authentication information, pass it to IgniteClient builder:

var cfg = new IgniteClientConfiguration("127.0.0.1:10800")
{
	Authenticator = new BasicAuthenticator
	{
		Username = "myUser",
		Password = "myPassword"
	}
};
IIgniteClient client = await IgniteClient.StartAsync(cfg);

Limitations

There are limitations to user types that can be used for such a mapping. Some limitations are common, and others are platform-specific due to the programming language used.

  • Only flat field structure is supported, meaning no nesting user objects. This is because Ignite tables, and therefore tuples have flat structure themselves;

  • Fields should be mapped to Ignite types;

  • All fields in user type should either be mapped to Table column or explicitly excluded;

  • All columns from Table should be mapped to some field in the user type;

  • .NET only: Any type (class, struct, record) is supported as long as all fields can be mapped to Ignite types;

Usage Examples

public class Account
{
  public long Id { get; set; }
  public long Balance { get; set; }

  [NotMapped]
  public Guid UnmappedId { get; set; }
}

SQL API

GridGain 9 is focused on SQL, and SQL API is the primary way to work with the data. You can read more about supported SQL statements in the SQL Reference section. Here is how you can send SQL requests:

IResultSet<IIgniteTuple> resultSet = await client.Sql.ExecuteAsync(transaction: null, "select name from tbl where id = ?", 42);
List<IIgniteTuple> rows = await resultSet.ToListAsync();
IIgniteTuple row = rows.Single();
Debug.Assert(row["name"] as string == "John Doe");

SQL Scripts

The default API executes SQL statements one at a time. If you want to execute large SQL statements, pass them to the executeScript() method. These statements will be executed in order.

string script =
    "CREATE TABLE IF NOT EXISTS Person (id int primary key, city_id int, name varchar, age int, company varchar);" +
    "INSERT INTO Person (1,3, 'John', 43, 'Sample')";

await Client.Sql.ExecuteScriptAsync(script);

Transactions

All table operations in GridGain 9 are transactional. You can provide an explicit transaction as a first argument of any Table and SQL API call. If you do not provide an explicit transaction, an implicit one will be created for every call.

Here is how you can provide a transaction explicitly:

var accounts = table.GetKeyValueView<long, Account>();
await accounts.PutAsync(transaction: null, 42, new Account(16_000));

await using ITransaction tx = await client.Transactions.BeginAsync();

(Account account, bool hasValue) = await accounts.GetAsync(tx, 42);
account = account with { Balance = account.Balance + 500 };

await accounts.PutAsync(tx, 42, account);

Debug.Assert((await accounts.GetAsync(tx, 42)).Value.Balance == 16_500);

await tx.RollbackAsync();

Debug.Assert((await accounts.GetAsync(null, 42)).Value.Balance == 16_000);

public record Account(decimal Balance);

Table API

To execute table operations on a specific table, you need to get a specific view of the table and use one of its methods. You can only create new tables by using SQL API.

When working with tables, you can use built-in Tuple type, which is a set of key-value pairs underneath, or map the data to your own types for a strongly-typed access. Here is how you can work with tables:

Getting a Table Instance

First, get an instance of the table. To obtain an instance of table, use the IgniteTables.table(String) method. You can also use IgniteTables.tables() method to list all existing tables.

var existingTables = await Client.Tables.GetTablesAsync();
var firstTable = existingTables[0];

var myTable = await Client.Tables.GetTableAsync("MY_TABLE");

Basic Table Operations

Once you’ve got a table you need to get a specific view to choose how you want to operate table records.

Binary Record View

A binary record view. It can be used to operate table tuples directly.

IRecordView<IIgniteTuple> view = table.RecordBinaryView;

IIgniteTuple fullRecord = new IgniteTuple
{
  ["id"] = 42,
  ["name"] = "John Doe"
};

await view.UpsertAsync(transaction: null, fullRecord);

IIgniteTuple keyRecord = new IgniteTuple { ["id"] = 42 };
(IIgniteTuple value, bool hasValue) = await view.GetAsync(transaction: null, keyRecord);

Debug.Assert(hasValue);
Debug.Assert(value.FieldCount == 2);
Debug.Assert(value["id"] as int? == 42);
Debug.Assert(value["name"] as string == "John Doe");

Record View

A record view mapped to a user type. It can be used to operate table using user objects which are mapped to table tuples.

var pocoView = table.GetRecordView<Poco>();

await pocoView.UpsertAsync(transaction: null, new Poco(42, "John Doe"));
var (value, hasValue) = await pocoView.GetAsync(transaction: null, new Poco(42));

Debug.Assert(hasValue);
Debug.Assert(value.Name == "John Doe");

public record Poco(long Id, string? Name = null);

Key-Value Binary View

A binary key-value view. It can be used to operate table using key and value tuples separately.

IKeyValueView<IIgniteTuple, IIgniteTuple> kvView = table.KeyValueBinaryView;

IIgniteTuple key = new IgniteTuple { ["id"] = 42 };
IIgniteTuple val = new IgniteTuple { ["name"] = "John Doe" };

await kvView.PutAsync(transaction: null, key, val);
(IIgniteTuple? value, bool hasValue) = await kvView.GetAsync(transaction: null, key);

Debug.Assert(hasValue);
Debug.Assert(value.FieldCount == 1);
Debug.Assert(value["name"] as string == "John Doe");

Key-Value View

A key-value view with user objects. It can be used to operate table using key and value user objects mapped to table tuples.

IKeyValueView<long, Poco> kvView = table.GetKeyValueView<long, Poco>();

await kvView.PutAsync(transaction: null, 42, new Poco(Id: 0, Name: "John Doe"));
(Poco? value, bool hasValue) = await kvView.GetAsync(transaction: null, 42);

Debug.Assert(hasValue);
Debug.Assert(value.Name == "John Doe");

public record Poco(long Id, string? Name = null);

Streaming Data

To stream a large amount of data, use the data streamer. Data streaming provides a quicker and more efficient way to load, organize and optimally distribute your data. Data streamer accepts a stream of data and distributes data entries across the cluster, where the processing takes place. Data streaming is available in all table views.

data streaming

Data streaming provides at-least-once delivery guarantee.

Using Data Streamer API

public async Task TestBasicStreamingRecordBinaryView()
{
    var options = DataStreamerOptions.Default with { BatchSize = 10 };
    var data = Enumerable.Range(0, Count).Select(x => new IgniteTuple { ["id"] = 1L, ["name"] = "foo" }).ToList();

    await TupleView.StreamDataAsync(data.ToAsyncEnumerable(), options);
}

Client Metrics

Metrics are exposed by the .NET client through the System.Diagnostics.Metrics API with the Apache.Ignite meter name. For example, here is how you can access GridGain metrics by using the dotnet-counters tool:

dotnet-counters monitor --counters Apache.Ignite,System.Runtime --process-id PID

You can also get metrics in your code by creating a listener:

var listener = new MeterListener();
listener.InstrumentPublished = (instrument, meterListener) =>
{
    if (instrument.Meter.Name == "Apache.Ignite")
    {
        meterListener.EnableMeasurementEvents(instrument);
    }
};
listener.SetMeasurementEventCallback<int>(
    (instrument, measurement, tags, state) => Console.WriteLine($"{instrument.Name}: {measurement}"));

listener.Start();

Available .NET Metrics

Metric name Description

connections-active

The number of currently active connections.

connections-established

The number of established connections.

connections-lost

The number of connections lost.

connections-lost-timeout

The number of connections lost due to a timeout.

handshakes-failed

The number of failed handshakes.

handshakes-failed-timeout

The number of handshakes that failed due to a timeout.

requests-active

The number of currently active requests.

requests-sent

The number of requests sent.

requests-completed

The number of completed requests. Requests are completed once a response is received.

requests-retried

The number of request retries.

requests-failed

The number of failed requests.

bytes-sent

The amount of bytes sent.

bytes-received

The amount of bytes received.

streamer-batches-sent

The number of data streamer batches sent.

streamer-items-sent

The number of data streamer items sent.

streamer-batches-active

The number of existing data streamer batches.

streamer-items-queued

The number of queued data streamer items.

© 2024 GridGain Systems, Inc. All Rights Reserved. Privacy Policy | Legal Notices. GridGain® is a registered trademark of GridGain Systems, Inc.
Apache, Apache Ignite, the Apache feather and the Apache Ignite logo are either registered trademarks or trademarks of The Apache Software Foundation.

Last updated on Jul 13, 2024