Log++ -- A Logging Framework for Modern Development

Log++ is a logging framework designed to support modern development needs.

• Log++ is fast! 4-7 times faster than bunyan and 2 times faster than pino. Log++ can process a log statement in under 1µs of blocking time, with the high cost formatting operations done as a background tasking using a N-API module.
• Log++ helps banish log spew with novel multi-level logging support. All (verbose) log messages are initially stored in an in-memory buffer for high performance and if they are needed for core dumps. However, these can be filitered out when writting so they don't pollute the saved logs with overly detailed information.
• Log++ logs are machine parsable. The logging output provides structured and machine parsable formats (even with printf style logging). These formats can be explicitly provided/managed via external JSON specifications. Common prefix data and macros for embedding useful system information are also available.
• Log++ provides modular flexible logging with unified control of logging levels, categories, and output across modules. Log++ supports both child loggers and logic for controlling the output from subloggers created by other modules that your application uses.

Basic Usage

Log++ is designed to be simple to use and provides a nearly drop in replacement for existing logging frameworks.

//Import the logpp module and create a logger (named "myapp") 
const log = require("logpp")("myapp");

//Add a format for better structured log output and provide it on the logger (prefixed with a "$")
log.addFormat("Hello", "Hello %s!!!");

//Emit the message specified by the format -- Hello "World"!!!
log.info(log.$Hello, "World");

//Or emit message given by a printf style format -- Hello "printf"!!!
log.info("Hello %s!!!", "printf");

Performance

Log++ is designed to minimize logging overhead on your application so you can use it to provide rich/useful information about your application without impacting your users. As a very rough benchmark we have a comparison with three other popular loggers -- Bunyan, Debug, and Pino (Extreme). The first 3 benchmarks are taken from Pino and the last one is from us. Each message is written 100k times and these timings are from an Intel Core i7-5600 running Node-V8 10.0 on Windows 10.

• Basic: info('hello world -- logger')
• String: info('hello %s', 'world')
• Multi: info('hello %s %j %d', 'world', { obj: true }, 4)
• Complex: info('hello at %j from %s with %j %n -- %s', new Date(), app, ['iter', { f: i, g: i.toString() }], i - 5, (i % 2 === 0 ? 'ok' : 'skip'))

The results in the table below show a representative timing taken for each logger framework on each benchmark and the speedup of Log++ over the next best performing framework.

Logger	Basic	String	Multi	Complex
Bunyan	690 ms	779 ms	1106 ms	1578 ms
Debug	266 ms	385 ms	515 ms	842 ms
Pino	141 ms	211 ms	378 ms	887 ms
Log++	71 ms	109 ms	253 ms	438 ms
Speedup	1.98x	1.93x	1.49x	1.92x

As seen in our results Log++ is consistently the lowest overhead logger, in most cases spending nearly 2x less time blocking the main event thread, than any of the others.

Logging Levels and Categories

Log++ supports 8 logging levels. At each level only messages at the given level and higher will be processed, any messages at lower levels will be nop'd or filtered (see multi-level logging below). For each level the logger exports a function with the corresponding name. Thus to log a message at the INFO level you would call log.info(...) and to log at the DEBUG level you can call log.debug(...). The available levels and order is:

• OFF        // all messages are disabled
• FATAL     // fatal messages only
• ERROR   // error and higher
• WARN    // warn and higher
• INFO      // info and higher -- default for emit (see multi-level logging)
• DETAIL   // detail and higher -- default for in-memory (see multi-level logging)
• DEBUG   // debug and higher
• TRACE    // trace and higher
• ALL         // all messages are enabled

In addition to adjusting the level at which log messages are processed Log++ also supports message categories to organize log messages and to selectively enable/disable them at a fine grain level. Categories can be defined using any name desired and enabled/disabled on a per logger basis.

Setting up and using a logging category can be done as follows:

//Define and enable the "performance" category
log.enableCategory("Performance", true);

//Emit the message specified by the format under the given category-- "Hello World!!!"
log.info(log.$$Performance, log.$Hello);

//Disable the "performance" category
logpp.enableCategory("Performance", false);

//NOP since Performance category is disabled now
log.info(log.$$Performance, logpp.$Hello);

Thus categories provide a simple way to selectively turn off/on related streams of logging data and (like logging levels) can be dynamically adjusted during execution.

Log++ also provides simple conditional logging with If versions of all the unconditional logging functions. These functions take a boolean as the first argument and only perform the logging work if the value is true.

log.info(log.$Hello); //"Hello World!!!"

let x = 1;
log.infoIf(x === 1, log.$Hello); //"Hello World!!!"
log.infoIf(x === 0, log.$Hello); //NOP

Multi-Level Logging

In the default mode Log++ operates with a novel multi-level logging system that consists of two layers and three phases.

1. An in memory buffer where all enabled messages are synchronously copied into during the logger call (info/debug/etc.). In this step the formatter info + information on the arguments is copied but no formatting or emitting is done.
2. Asynchronously messages + data from the in memory buffer that are over time or memory limits (see Multi-Level Log Layer Management below) are checked and either discarded or copied to a format & emit buffer. The decision to copy/discard is governed by the level the message was logged at and the emit level the loggers are running at.
3. In a native module, running on a background thread, the messages to be emitted are formatted into specified wire format (currently JSON and Binary -- see Formatters below). When this is complete the results can be written directly to stdout or a user provided writable stream (see Emit Channels below).

With this design the only blocking cost on the main thread for a log operation is an enabled check, an immediate check/copy of the message data, and a bulk filter copy. This time is much smaller than the format and emit time which then happens on an asynchronous background thread -- resulting in a 5x or more reduction in performance overhead when compared to console.log style logging.

In addition to this default asynchronous mode Log++ also supports completely synchronous processing (at lower performance) as an option and as the default when the debugger is attached vs. high performance async when running in production (see Sync vs. Async below).

Beyond the performance implications of splitting the logging into 2 layers there are two other substantial benefits:

1. Since the initial logging of a message is cheap (~1µs) it is feasible to log at a much higher rate than previously possible, say DETAIL or DEBUG instead of INFO or just WARN. Then, when an error occurs this invaluable information is present in the in-memory buffer and can be included with the rest of the error dump.
2. While this high fidelity logging data is desireable when analyzing an error or performance issue for normal execution it is simply clutter that unnecessarily takes up and space and complicates other informational log analysis tasks. The ability to discard messages as they move to the emit buffer allows us to eliminate this clutter and only save messages that are generally useful and worth longer term storage (say INFO or just WARN).

Full Log Write Hooks

To maximize the value of the in-memory log ability we support the hooking of common failure events and will perform an explicit flush of the full in-memory log. The auto-hooked events are:

• Unhandled exception
• Exit with a non-zero exit code

In other cases you may want to programatically (and synchronously) produce a full dump of the in-memory log. This can be done with the emitLogSync(FULL, DETAIL) API which synchronously returns the formatted log data as a string (where the log can be fully flushed and can include details from all the log messages without any filtering).

Multi-Level Log Layer Management

Log++ provides a range of settings to control how data is stored and flows between the levels in the log and is ultimately emitted.

Sync vs. Async: Log++ supports asynchronous logging, for best performance and other benefits of multi-level logging, as well as fully synchronous logging for simpler debugging and testing scenarios. The mode can be set with the flushMode option -- either 'ASYNC' | 'SYNC' | 'NOP'.

The default mode is 'ASYNC' which enables formatting in the background of messages that have been added to the emit worklist. This option provides the best performance. The 'SYNC' mode does synchronous processing of the messages in the emit worklist, resulting in higher overhead than the 'ASYNC' mode but, if used in combination with a small flushCount and time/space constraints on the in-memory buffer provides a fast and consistent logging setup. Finally, the 'NOP' option prevents any implicit processing of messages in the emit list and let's the user control when they are written (e.g., using emitLogSync).

Flush Heuristics: Log++ provides several parameters to control how much data is stored in the in-memory log and how aggressively it is flushed to the emit worklist.

To control how much space the in-memory list consumes Log++ provides two parameters:

• Age -- since we want to keep details available in the event of an error or other diagnostic need. Any message older than the limit is filtered/processed as needed.
• Size -- since there may be unusual bursts of logging, and we don't want log data to excessively displace user info, we also support a maximum number of entires (values) in the in-memory log as a failstop. Log++ works to keep the number of slots used under this amount.

By default Log++ writes the final formatted results directly into stdout. However, for applications may also want to send the data to a file or a remote server. This can be done by setting the flushTarget option as stream and providing the writable stream as the stream option value.

Message Formats and Formatters

The primary mechanism for writing messages in Log++ is to use structured formatters. By separating the definition of formats from their use we are able to (1) optimize a format which will be repeatedly used and (2) generate a more uniform log structure that can be automatically parsed by later analysis tools.

Message Format Structure

The message formats can use either classic printf style format strings or special JSON style messages. There are 2 classes of format specifiers that are permitted in these message formats -- expando macros for common environmental data, like timestamps, source lines, or request ids and traditional value formats for numbers/strings as well as object formats for JSON style object/array value formatting.

Expando Macros:

• #host -- name of the host
• #app -- name of the root app
• #logger -- name of the logger
• #source -- source location of log statment (file, line)
• #wallclock -- wallclock timestamp (iso format)
• #timestamp -- logical timestamp -- incremented on each use and on explicit advances
• #callback -- the current callback id
• #request -- the current request id (for http requests)
• ## -- a literal # Some example uses of these macros in format messages include:

log.addFormat("Clock", "current time is #wallclock");
log.addFormat("CallbackId", "current callback is #callback");
log.addFormat("Timestamp", "timestamp #timestamp");

log.info(log.$Clock); //current time is "2018-05-08T05:29:55.0512Z"
log.info(log.$CallbackId); //current callback is 1

log.info(log.$Timestamp); //timestamp 0
log.info(log.$Timestamp); //timestamp 1

Value & Object Formats:

• %b -- a boolean value
• %n -- a number
• %s -- a string
• %di -- a date formatted as iso time format
• %dl -- a date formatted as local time format
• %j -- a JavaScript value where objects/arrays are expanded up to 2 levels and 32 items in any level
• %j<d,l> -- a JavaScript value where objects/arrays are expanded up to d levels (default is 2 and '' is unlimited) and l items in any level (default is 32 and '' is unlimited)
• %% -- a literal %

For objects and arrays the format allows for depth and length limits on the expansion. So, a specifier could be %j (default expansion), %j<1,> (max depth of 1 default length), or %j<2,10> (max depth 2 max length 10). Cycle detection is done by default and noted with a special "<Cycle>" token in the output. For variables that are not naturally formattable are printed as "<OpaqueValue>".

Some example uses of these in format messages include:

log.addFormat("Number", "A number %n");
log.addFormat("DateLocal", "A date %dl");
log.addFormat("Object", "An object %j");
log.addFormat("ObjectWDepth", "A shallow object %j<1,>");

log.info(log.$Number, 2); // A number 2
log.info(log.$DateLocal, new Date()); // A date "Mon May 07 2018 22:29:55 GMT-0700 (Pacific Daylight Time)"
log.info(log.$Object, {f: 3, g: [1, 2]}); // An object {"f": 3, "g": [1, 2]}
log.info(log.$ObjectWDepth, {f: 3, g: [1, 2]}); // A shallow object {"f": 3, "g": "..."}

The general (j) format is a catchall that will format a value using the default options for whatever the type of the value is.

log.addFormat("General", "Value is %j");

log.info(log.$General, 2); // Value is 2
log.info(log.$General, "ok"); // Value is "ok"
log.info(log.$General, {f: 3}); // Value is {"f": 3}

JSON Message Formats: In addition to classic printf style formats Log++ also supports hybrid JSON message formats. In these formats you can place a format specifier in any value position. For example:

log.addFormat("Json", {kind: "start", time: "#wallclock", value: "%j"});
log.info(log.$Json, [1, 2]); // {"kind": "start", "time": "2018-05-08T05:29:55.0512Z", "value": [1, 2]}

Standard Prefixs: To simplify logging Log++ provides a simple standard prefix option that is prepended to every message during formatting of the following form:

LEVEL#CATEGORY @ TIME from HOST::LOGGER_NAME |

This metadata is frequently useful, can be efficiently provided automatically, and the builtin support simplifies the job of the logging format specification.

Format registration: Formats can be registered in a number of ways:

• Single programatic registration using the addFormat method.
• Bulk registration using the addFormats method which takes a JSON object where each property is a format name and value is a format specifier or a string which refers to a JSON file to load the format object from.
• As the formats configuration option -- provided as a JSON object or file name to load from.

Emit Formats

A common representation for logging data is JSON or newline separated messages. By default Log++ uses newline separated UTF8 text with and optional prefix followed by the payload contents as specified by the formatter.

[In Progress] Log++ also supports an efficient and compact binary format which is faster to process and more space efficient to transport/store. This format can be post-processed into human readable text using via the provided command line utility --humanify.

Managing Loggers

A key challenge in logging in an ecosystem such as Node.js is the heavy use of modules and components. Ideally all of the logging in these frameworks should coordinate the levels of active logging, the sink of these logs, and merging the message streams. Additionally, the "master" application should be able to disable or reduce rates of logging from submodules that are not of interest. To support this we introduce the concept of a root logger which is able to control all the sublogger actions (and subloggers cannot override these settings). Log++ also supports the frequently useful child logger scenario.

Root Logger and SubLoggers

The root logger is the logger created in the require.main.file module (i.e., the first loaded file). Each logger is created with a module name and all loggers created with the same name share the same logger. The root logger can set the emit level, sink, etc. and enable/disable or set levels of subloggers explicitly. If not explicitly set subloggers are restricted to emitting at the WARN level. All messages are merged and managed automatically according to the settings of the root logger. Consider the following example:

/*app.js*/
const log = require("logpp")("main");
log.addFormat("Hello", "Hello World!!!");

const helper = require("./helper");
helper.doit(); //Emit "Msg2" from sublogger

log.info(log.$Hello); //Emit "Hello World!!!" since we are root logger

/*helper.js*/
const log = require("logpp")("sub");

log.addFormat("Msg1", "Msg1");
log.addFormat("Msg2", "Msg2");

function doit() {
    log.setLoggingLevel(log.Levels.DETAIL); //NOP -- not root logger
    log.info(log.$Msg1); //NOP -- subloggers are set to WARN level by default

    log.warn(log.$Msg2); //Processed
}

module.exports.doit = doit;

In general the default, of allowing subloggers to emit WARN or higher, provides a balance where critical information is logged but (likely) irrelevant logging output is suppressed. However, in cases where the logging output from a module is of interest, you can explicitly set a sublogger configuration from the root logger using the setSubLoggerLevel API:

/*app2.js*/
const log = require("logpp")("main");
log.setSubLoggerLevel("sub", log.Levels.INFO); //configure sublogger

const helper = require("./helper");
helper.doit(); //Emit "Msg1" and "Msg2" from sublogger

In cases where there is a module with a sublogger that is completely uninteresting or that is very noisy you can completely diable the sublogger using the disableSubLogger API as well.

Child Loggers

Child loggers provide a simple way to specialize a logger for a particular section of an application with a default block of information that is output with each log message. For exmaple:

const log = require("logpp")("main");
log.addFormat("Hello", "Hello %s!!!");

function doit(v) {
  const childlog = log.childLogger({arg: v});
  childlog.info(childlog.$Hello, v);
}

doit("Ok"); //Emit {arg:"Ok"} -- "Hello Ok!!!" from child logger 

Finally, child loggers can stack -- a child logger may create another child logger. In this case we will extend the child logger value with the newly provided info.

API Specification

require("logpp")(NAME, [OPTIONS])

NAME - string that is the name of the logger. If the same name is used in multiple invocations the same logger will be returned.
OPTIONS - an object where each property is a configuration option for the created logger.

• memoryLevel - string name of enabled level for in-memory buffer (default "DETAIL").
• emitLevel - string name of enabled level for formatting and emitting (default "INFO").
• defaultSubloggerLevel - string name of level that loggers in submodules memoryLevels are forced to (default "WARN").
• flushCount - number of log messages are added to the in-memory log before attempting to process them (default 64).
• flushTarget - the target output of the processed emit log data "console"|"stream" (default "console").
• flushMode - how messages are processed for emit "SYNC"|"ASYNC"|"NOP" (default "ASYNC").
• flushCallback - NOT SUPPORTED YET
• prefix - boolean specifying if default prefix is included in all emitted messages (default true).
• bufferSizeLimit -- in-memory buffer size threshold for processing, messages may not be flushed if under this limit (default 1024 ~ 16kb).
• bufferTimeLimit -- in-memory age threshold for processing, messages may not be flushed if younger than this limit (default 500ms).
• formats -- JSON object or file name to load formats from (default empty).
• categories -- provided as a JSON object or file name to load category definitions from (default empty).
• subloggers -- provided as a JSON object or file name to load sublogger configurations from (default empty).

The first logger created in the file that matches require.main.filename will be the root logger. Loggers created in modules before the root logger is defined are OFF and will be set to their defaults if/when the root logger is defined. Loggers created after the root logger is defined will have their levels/behavior set according to the root logger configurations and/or defaults.

As buffered/delayed logging output can be confusing during interactive debugging Log++ checks for launch with --inspect and if detected sets the default flushCount = 0 and flushMode = "SYNC" so that all log messages are immediately processed and output.

this.setLoggingLevel(LEVEL)

LEVEL - the desired level to set for in-memory processing.

Each logger has these values accessible on this.Levels.LEVEL (e.g., log.Levels.INFO).

this.setEmitLevel(LEVEL)

LEVEL - the desired level to set for formatting and emit.

Each logger has these values accessible on this.Levels.LEVEL (e.g., log.Levels.INFO).

this.enableCategory(NAME, ENABLED)

NAME - the string name of the category to configure. ENABLED - the boolean enabled value for the category.

Each logger has these values accessible on this.$$NAME (e.g., log.$$Performance).

this.enableCategories(ARG)

ARG a JSON object or string filename with JSON object where each property is a category name and each value is the enabled value.

this.addFormat(NAME, FORMAT)

NAME the string name of the format. FORMAT string sprintf format or JSON Object/Array format object.

Each logger has the formats accessible on this.$NAME (e.g., log.$Hello).

this.addFormats(ARG)

ARG a JSON object or string filename with JSON object where each property is a format name and each value is the format value.

this.setMsgTimeLimit(LIMIT)

LIMIT the age limit in ms that governs when messages are removed from, and processed if needed, the in-memory log.

this.setMsgSpaceLimit(LIMIT)

LIMIT the space limit in slots (1 slot ~16bytes) that governs when messages are removed from, and processed if needed, the in-memory log.

LOG_FUNCTION(FORMAT, ...ARGS) and LOG_FUNCTION(CATEGORY, FORMAT, ...ARGS)

LOG_FUNCTION - a log level function fatal | error | warn | info | detail | debug | trace
CATEGORY - (optional) the desired category to process this log call with. Each logger has these values accessible as names prefixed with $$ (e.g., log.$$CATEGORY).
FORMAT - the format to use in generating this message. Multiple format specifications are possible:

• Pre-Defined: Using a format name added previously -- e.g. log.info(log.$Hello, "World")
• Explicit String: Creating and using an ad-hoc format string -- e.g. log.info("Hello %s", "World")
• Explicit Object or Array: Simple JSON formatting of an Object or Array (other values not supported) -- e.g. log.info({f: 3, ok: true})

ARGS - the rest of the arguments that are needed by the format specifier.

LOG_FUNCTION_COND(COND, FORMAT, ...ARGS) and LOG_FUNCTION_COND(COND, CATEGORY, FORMAT, ...ARGS)

LOG_FUNCTION_COND - a conditional log level function fatalIf | errorIf | warnIf | infoIf | detailIf | debugIf | traceIf
COND a boolean that, if true, the message will be processed and, if false, is a nop. \

The other args are the same as for the unconditional log method.

this.emitLogSync(FULL, DETAIL)

FULL - true if all messages should be flushed and false if only those over the age/size limit are eligible for processing
DETAIL - true if all messages should be processed regardless of level and false if filtering should be applied as usual.

Common uses include:

• this.emitLogSync(true, true) all messages are flushed -- good for panic output
• this.emitLogSync(true, false) all messages are flushed but filtered -- good for action completed want to drain log
• this.emitLogSync(true, true) partial flush with filter -- maybe useful to keep memory use down from buffering?

this.setSubLoggerLevel(SUBLOGGER_NAME, LEVEL)

SUBLOGGER_NAME string name of the sublogger to change the emit level on. LEVEL the new emit level for the sublogger.

this.disableSubLogger(SUBLOGGER_NAME)

SUBLOGGER_NAME string name of the sublogger to disable -- no output will be generated.

this.configureSubloggers(ARG)

ARG a JSON object or string filename with JSON object with 2 properties -- enabled which is a JSON object where each property is a sublogger name and each value is the enabled value and disabled which is a JSON array of names of diabled subloggers.

this.childLogger(PREFIX_DATA)

PREFIX_DATA a JSON Object that is the prefix data to be associated with all messages emitted from the child logger. If there was a previous prefix value this extends it.