logmet-client

This is a nodejs module for sending data to and querying data from Logmet. This module defines two classes: LogmetProducer and LogmetConsumer. This document describes these two classes and provides some guidance on how to use them to, respectively, send and query data.

License

This library is released under the Apache 2.0 license.

Contributing

Contributions are welcome via pull requests. Please submit your pull request against the Developer's Certificate of Origin, adding a line like the following to your commit messages, using your name and e-mail address:

Signed-off-by: John Doe [email protected]

LogmetProducer Class

This class can be used to send nodejs objects to Logmet. It translates the objects into a format that allows the object structure to be preserved on Logmet so that the data becomes queriable. It is important to note that this class must be treated as a singleton, that is, in one nodejs program it makes sense to have only one instance of this class.

The constructor LogmetProducer is defined as follows:

function LogmetProducer(endpoint, port, tenantOrSupertenantId, logmetToken, isSuperTenant, options)

The above constructor takes the following parameters:

• logmetEndpoint: The Logmet host to which the data will be sent. For example, if the target is the production environment of the US-South datacenter, the value of this parameter should be logs.opvis.bluemix.net.
• logmetPort: The port Logmet to which Logmet clients should connect for sending data.
• logmetTenant: The value of this parameter represents either a Bluemix space id (tenant id), or the id of a Logmet supertenant.
• logmetToken: This is the Logmet token associated with the chosen logmetTenant. This token can be obtained by querying Logmet. It never expires.
• isSuperTenant: This is a Boolean-valued parameter indicating whether or not the value passed to logmetTenant represents a supertenant.
• options: This object allows you to optionally override defaults in the client. Currently, the only supported override is for bufferSize

Usage

Below is a sample code showing how to use the LogmetProducer class. In the example, we assume that the logmet-client module code is one level up in the directory hierarchy relative to the sample code.

var logmet = require('../logmet-client');

var logmetEndpoint = 'logs.opvis.bluemix.net';
var logmetPort = 9091;
var logmetTenant = process.env.LOGMET_TENANT;
var logmetToken = process.env.LOGMET_TOKEN;

var logmetProducer = new logmet.LogmetProducer(logmetEndpoint, logmetPort, logmetTenant, logmetToken, false, {bufferSize: 100});

logmetProducer.connect(function(error, status) {
    if (error) {
        console.log('Connection with Logmet failed. ERROR: ' + error);
    } else if (status.handshakeCompleted) {
        console.log('LogmetClient is ready to send data.');
    }
});

function sendData(event) {
    logmetProducer.sendData(event, 'tool_id', logmetTenant, function(error, status) {
        if (error) {
            console.log('Logmet client rejected the data. ERROR: ' + error);
        } else {
            console.log('Logmet client accepted the data.');
            if (!status.connectionActive) {
                console.log('WARNING: Logmet client not connected to Logmet, data waiting in buffer.')
            }
        }
    });
}

Before calling the sendData function for the first time, the program should call the function connect. This function will establish a persistent connection with Logmet. The connect function takes a callback as its only argument. Upon successfully connecting to Logmet, connect will pass to the callback a status object with the Boolean-valued field handshakeCompleted set to true. If the sendData function is called before the connect function callback returns, incoming data will be placed in the buffer and will be sent to Logmet once a connection with Logmet is established.

The sendData function, as shown in the sample above, takes the following parameters in that order:

• the object to be send to Logmet.
• the Elasticsearch type to be associated with the object.
• the Bluemix space id corresponding to the owner of the data. If the constructor was called with a regular tenant id, that is, isSuperTenant was set to false, then the value of this parameter must match the id given to the constructor. Differently, if the constructor was called with isSuperTenant set to true, then the value of this parameter will contain a Bluemix space id corresponding to the tenant who will own the data, on behalf of whom the supertenant is sending the data.
• A callback function, indicating whether the data was accepted or not.

If the data buffer is full, the data will not be accepted and the callback function will receive an error message in the error argument. The data returned by the callback function in the status argument is an object containing a Boolean-valued field:

{
  connectionActive: false
}

The connectionActive field indicates whether the logmet-client is currently connected to Logmet. When the connection is inactive, data is placed in the buffer but won't be sent to Logmet until a connection is established.

LogmetConsumer Class

This class can be used to query Logmet for data. This is by no means the only way to query Logmet for data stored with the LogmetProducer class. It is just a convenient way to abstract a few details such as setting Logmet HTTP headers and dealing with the Logmet multitenancy-based index-naming approach.

The constructor LogmetConsumer is defined as follows:

function LogmetConsumer(logmetQueryEndpoint)

The above constructor takes one argument, namely, logmetQueryEndpoint. The value of that argument must be the host name exposed by Logmet for querying purposes. For example, if the target is the production environment of the US-South datacenter, the value of this parameter should be logmet.ng.bluemix.net.

Usage

Below is a sample code showing how to use the LogmetConsumer class. In the example, we assume that the logmet-client module code is one level up in the directory hierarchy relative to the sample code.

var logmet = require('../logmet-client');

var logmetQueryEndpoint = 'logmet.ng.bluemix.net';

var logmetConsumer = new logmet.LogmetConsumer(logmetQueryEndpoint);

var queryDSLBody = {
		query: {
			match: {
				id: "fabioPipeline1"
			}
        }
};

// Omitted initializations of tenantId and bearerToken

logmetConsumer.query(tenantId, bearerToken, 'tool_id', queryDSLBody, function(error, documents) {
	if (error != '') {
		console.log('Query returned an error: ' + error);
	} else {
		console.log('Documents returned by Logmet:');
		console.log(documents);
	}
});

The query function, used in the sample above, takes the following parameters in that order:

• The id of the tenant (Bluemix space id) who owns the data that is being queried.
• A valid Bluemix bearer token belonging to the identified tenant.
• The Elasticsearch type that has been associated with the data being queried.
• A query expressed in the Elasticsearch query DSL codified as a nodejs object.
• A callback function used to process errors as well as the actual data returned by the query.

In case of error, the callback function will receive the error message in the error argument. All documents retrieved by the query will be passed to the callback function as an array of objects assigned to the argument documents.

Debugging

The default logging level for the library is warn, to customize it set the environment variable logmet_client_njs_level to your desired logging level. Example: export logmet_client_njs_level='debug' && node index. Other valid logging options are trace, info, warn, error and fatal.