Introduction

This module is able to trace the execution of the following services:

• data export from SQL query

• web service remote invocation

• enqueuing and dequeuing of elements from queues

• web services execution and by and large any server-side javascript action

Platform can automatically trace these events:

• out of memory

• authentication failure

• javascript syntax errors

• javascript execution errors

• database locks

• data tracking over web service remote invocations

• input/output data (HTTP request parameters, headers, body, response)

• execution outcome and errors

• distributed transaction id over remove ws invocations

• internal state

• execution start/end datetime

• current elaboration state (started, ended, interrupted)

• tagging some of the input data (automatically retrieved from the input, up to 10); these tag values can be used later to search for logged data, filtered by these tags

Programmatically, a Platform developer can inject additional events, through server-side javascript utility methods:

• messages to append to the server-side javascript action execution (a message has a message type having one of the following values: FINEST, DEBUG, INFO, ERROR, FATAL)

• files to read/write and their state (in progress, completed)

• the execution state and outcome

Thanks to these features, it is possible to define any number of monitored services, each identified by a service code and automate logging data and behavior. Optionally, a Platform developer can include additional invocations, to fine tune the logging and tracing of the service.

Once configured a series of services to monitor, Platform will start gathering data about their execution, in terms of elaborations, error messages (and my and large any kind of messages), processed files (if there is any).

Finally, it is possible to search for elaborations and messages/files.

Moreover, a notification sub-system is available, in order to automate the email notification of some events (errors), which can be configured at service level: according to the message type (e.g. only ERROR or FATAL type messages), it is possible to notify the event to a specific set of users via email.