Communicate.Archeo 1.2.5

Usage

This package is used to easily connect your C# application to the Archeo logging application.

To start logging to Archeo you need to specify what happens if you loose your connection to Archeo. We specify this by implementing the IBackupLogger interface. Logging failures can happen in a multitude of scenarios and you should implement alerting on the backup logger to make your organization aware of potential failures. A simple backup logger implementation could look something like this if you use Application Insights as the backup logger reciever: (Any backup logging solution could be implemented, AI is only used as an example.)

    public class BackupLogger : IBackupLogger
    {
        private string AiKey { get; set; }

        private TelemetryClient telemetryClient;

        public BackupLogger()
        {
            AiKey = Environment.GetEnvironmentVariable("APPINSIGHTS_INSTRUMENTATIONKEY", EnvironmentVariableTarget.Process);
            telemetryClient = new TelemetryClient() { InstrumentationKey = AiKey };
        }

        public void LogInformation(string log)
        {
            var evt = new EventTelemetry(log);
            telemetryClient.TrackEvent(evt);
        }

        public void LogError(string message, List<LogStep> logSteps)
        {
            telemetryClient.TrackTrace(message, new Dictionary<string, string>() { { "Logstep object", JsonConvert.SerializeObject(logSteps) } });
        }

        public void LogException(Exception e, List<LogStep> logSteps)
        {
            telemetryClient.TrackException(e, new Dictionary<string, string>() { { "Logstep object", JsonConvert.SerializeObject(logSteps) } });
        }
    }

If you use application insights the team has released a NuGet package with a backup logger implementation for convenience: Communicate.Archeo.BackupLogger.ApplicationInsights

Creating the log object

The Archeo SDK log object supports Dependency Injection as a scoped service only. It's important that you ensure that the object lives only once per web request and not as a singleton.

Using HttpClientFactory

In your startup.cs you need to declare a HttpClient. You can name it whatever you want but a default name is supplied in the ArcheoDefaults constants class. The Uri for Archeo is also supplied in ArcheoDefaults as seen below:

            services.AddHttpClient(ArcheoDefaults.HttpClientName, client =>
            {
                client.BaseAddress = new Uri(ArcheoDefaults.ArcheoEndpoint);
                client.DefaultRequestHeaders.Add("APIKEY", GetAppSetting("ArcheoApiKey"));
            });

You also need to add the IArcheoLogger itself as a scoped service. You can use dependency injection to insert both the IHttpFactory and the IBackupLogger or you can supply them explicitly:

      services.AddScoped<IArcheoLogger>(sp => new ArcheoLogger(sp.GetService<IHttpClientFactory>(), new BackupLogger()));
      Or
      services.AddScoped<IBackupLogger, BackupLogger>();
      services.AddScoped<IArcheoLogger, ArcheoLogger>();

Letting the SDK supply the HttpClient

If you don't want to have a relationship with the HttpClient used for the sending of logs you can use the constructor for ArcheoLogger like this:

            IExtendedBackupLogger backupLogger = new AiExtendedBackupLogger(Settings.GetAppSetting("APPINSIGHTS_INSTRUMENTATIONKEY"));
            ArcheoLogger logger = new ArcheoLogger(backupLogger, new ArcheoConfiguration() { ApiKey = Settings.GetAppSetting("ArcheoApiKey") });

Here we also use the AiExtended logger which supplies extra methods for contacting the backup logger.

Supplying the HttpClient yourself

If you don't want to have a relationship with the HttpClient used for the sending of logs you can use the constructor for ArcheoLogger like this:

            IExtendedBackupLogger backupLogger = new AiExtendedBackupLogger(Settings.GetAppSetting("APPINSIGHTS_INSTRUMENTATIONKEY"));
            HttpClient client = new HttpClient();
            client.BaseAddress = new Uri("http://google.com");
            client.DefaultRequestHeaders.Add("APIKEY", "asd");

            ArcheoLogger logger = new ArcheoLogger(client, backupLogger);

Setting up the logger

To avoid having to send every parameter every time you use the logger this nuget provides the Arche.LogBase object. This object is used if parameters supplied on logging are null. The LogBase should be instantiated as early in the pipeline as possible. You simply set the object like this:

            logger.LogBase = new ArcheoLogBase()
            {
                MessageType = messageType,
                Reciever = partyConfig.ReceiverId,
                Sender = partyConfig.SenderId,
                TransactionId = transactionId,
                TransactionTag = transactionTag,
                TransactionType = partyConfig.DocumentType
            };

Logging

If you set up the LogBase you only need 2-3 parameters when logging. If you need to override any of the LogBase params on only a single log entry just supply the parameter. This nuget provides 4 logging methods as of now. Please see the code comments for details. Here are some simple examples that only work if LogBase is used:

archeo.LogException((Exception)e, $"Exception occured in {nameof(test)}");
archeo.LogHttpFailure((HttpResponseMessage)result, "Sending file to APIM failed");
archeo.Log(encoding.GetBytes(fileContent), "File successfully downloaded", $"file.file");
archeo.Log(fileContent, "Something failed")

Sending logs

After logging you share of logs you need to call await archeo.SendLogs(). This is usually done in the finallyclause in the catch all, but you are free to do this whenever you want. When logs are sent the list of logged steps is cleared but the LogBase is kept in case you want to log more withing the same scope. If you want to clear both objects you can call IArcheoLogger.Clear()