1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
|
package logging
import (
"fmt"
"io"
"os"
"time"
)
// Logger provides a standardised logging interface.
type Logger interface {
// SetLevel sets the minimum log level.
SetLevel(Level)
// Good returns true if the Logger is healthy.
Good() bool
// Status returns an error corresponding to the logger's state;
// if it's healthy (e.g. Good() returns true), Error will
// return nil.
Status() error
// Close gives the Logger the opportunity to perform any cleanup.
Close()
// Log messages consist of four components:
//
// 1. The **level** attaches a notion of priority to the log message.
// Several log levels are available:
//
// + FATAL (32): the system is in an unsuable state, and cannot
// continue to run. Most of the logging for this will cause the
// program to exit with an error code.
// + CRITICAL (16): critical conditions. The error, if uncorrected, is
// likely to cause a fatal condition shortly. An example is running
// out of disk space. This is something that the ops team should get
// paged for.
// + ERROR (8): error conditions. A single error doesn't require an
// ops team to be paged, but repeated errors should often trigger a
// page based on threshold triggers. An example is a network
// failure: it might be a transient failure (these do happen), but
// most of the time it's self-correcting.
// + WARNING (4): warning conditions. An example of this is a bad
// request sent to a server. This isn't an error on the part of the
// program, but it may be indicative of other things. Like errors,
// the ops team shouldn't be paged for errors, but a page might be
// triggered if a certain threshold of warnings is reached (which is
// typically much higher than errors). For example, repeated
// warnings might be a sign that the system is under attack.
// + INFO (2): informational message. This is a normal log message
// that is used to deliver information, such as recording
// requests. Ops teams are never paged for informational
// messages. This is the default log level.
// + DEBUG (1): debug-level message. These are only used during
// development or if a deployed system repeatedly sees abnormal
// errors.
//
// The numeric values indicate the priority of a given level.
//
// 2. The **actor** is used to specify which component is generating
// the log message. This could be the program name, or it could be
// a specific component inside the system.
//
// 3. The **event** is a short message indicating what happened. This is
// most like the traditional log message.
//
// 4. The **attributes** are an optional set of key-value string pairs that
// provide additional information.
//
// Additionally, each log message has an associated timestamp. For the
// text-based logs, this is "%FT%T%z"; for the binary logs, this is a
// 64-bit Unix timestamp. An example text-based timestamp might look like ::
//
// [2016-03-27T20:59:27-0700] [INFO] [actor:server event:request received] client=192.168.2.5 request-size=839
//
// Note that this is organised in a manner that facilitates parsing::
//
// /\[(\d{4}-\d{3}-\d{2}T\d{2}:\d{2}:\d{2}[+-]\d{4})\] \[(\w+\)]\) \[actor:(.+?) event:(.+?)\]/
//
// will cover the header:
//
// + ``$1`` contains the timestamp
// + ``$2`` contains the level
// + ``$3`` contains the actor
// + ``$4`` contains the event
Debug(actor, event string, attrs map[string]string)
Info(actor, event string, attrs map[string]string)
Warn(actor, event string, attrs map[string]string)
Error(actor, event string, attrs map[string]string)
Critical(actor, event string, attrs map[string]string)
Fatal(actor, event string, attrs map[string]string)
FatalCode(exitcode int, actor, event string, attrs map[string]string)
FatalNoDie(actor, event string, attrs map[string]string)
}
// A LogWriter is a Logger that operates on an io.Writer.
type LogWriter struct {
wo, we io.Writer
lvl Level
state error
snl bool // suppress newline
}
// NewLogWriter takes an output writer (wo) and an error writer (we),
// and produces a new Logger. If the error writer is nil, error logs
// will be multiplexed onto the output writer.
func NewLogWriter(wo, we io.Writer) *LogWriter {
if we == nil {
we = wo
}
return &LogWriter{
wo: wo,
we: we,
lvl: DefaultLevel,
state: nil,
}
}
func (lw *LogWriter) output(w io.Writer, lvl Level, actor, event string, attrs map[string]string) {
t := time.Now().Format(DateFormat)
fmt.Fprintf(w, "[%s] [%s] [actor:%s event:%s]", t, levelPrefix[lvl], actor, event)
for k, v := range attrs {
fmt.Fprintf(w, " %s=%s", k, v)
}
if !lw.snl {
fmt.Fprintf(w, "\n")
}
}
// Debug emits a debug-level message. These are only used during
// development or if a deployed system repeatedly sees abnormal
// errors.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Debug(actor, event string, attrs map[string]string) {
if lw.lvl > LevelDebug {
return
}
lw.output(lw.wo, LevelDebug, actor, event, attrs)
}
// Info emits an informational message. This is a normal log message
// that is used to deliver information, such as recording
// requests. Ops teams are never paged for informational
// messages. This is the default log level.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Info(actor, event string, attrs map[string]string) {
if lw.lvl > LevelInfo {
return
}
lw.output(lw.wo, LevelInfo, actor, event, attrs)
}
// Warn emits a warning message. An example of this is a bad request
// sent to a server. This isn't an error on the part of the program,
// but it may be indicative of other things. Like errors, the ops team
// shouldn't be paged for errors, but a page might be triggered if a
// certain threshold of warnings is reached (which is typically much
// higher than errors). For example, repeated warnings might be a sign
// that the system is under attack.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Warn(actor, event string, attrs map[string]string) {
if lw.lvl > LevelWarning {
return
}
lw.output(lw.we, LevelWarning, actor, event, attrs)
}
// Error emits an error message. A single error doesn't require an ops
// team to be paged, but repeated errors should often trigger a page
// based on threshold triggers. An example is a network failure: it
// might be a transient failure (these do happen), but most of the
// time it's self-correcting.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Error(actor, event string, attrs map[string]string) {
if lw.lvl > LevelError {
return
}
lw.output(lw.we, LevelError, actor, event, attrs)
}
// Critical emits a message indicating a critical condition. The
// error, if uncorrected, is likely to cause a fatal condition
// shortly. An example is running out of disk space. This is
// something that the ops team should get paged for.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Critical(actor, event string, attrs map[string]string) {
if lw.lvl > LevelCritical {
return
}
lw.output(lw.we, LevelCritical, actor, event, attrs)
}
// Fatal emits a message indicating that the system is in an unsuable
// state, and cannot continue to run. The program will exit with exit
// code 1.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) Fatal(actor, event string, attrs map[string]string) {
if lw.lvl > LevelFatal {
return
}
lw.output(lw.we, LevelFatal, actor, event, attrs)
os.Exit(1)
}
// Fatal emits a message indicating that the system is in an unsuable
// state, and cannot continue to run. The program will exit with the
// exit code speicfied in the exitcode argument.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) FatalCode(exitcode int, actor, event string, attrs map[string]string) {
if lw.lvl > LevelFatal {
return
}
lw.output(lw.we, LevelFatal, actor, event, attrs)
os.Exit(exitcode)
}
// Fatal emits a message indicating that the system is in an unsuable
// state, and cannot continue to run. The program will not exit; it is
// assumed that the caller has some final clean up to perform.
//
// Actor specifies the component emitting the message; event indicates
// the event that caused the log message to be emitted. attrs is a map
// of key-value string pairs that can be used to provide additional
// information.
func (lw *LogWriter) FatalNoDie(actor, event string, attrs map[string]string) {
if lw.lvl > LevelFatal {
return
}
lw.output(lw.we, LevelFatal, actor, event, attrs)
}
// Good returns true if the logger is healthy.
func (lw *LogWriter) Good() bool {
return lw.state == nil
}
// Status returns an error value from the logger if it isn't healthy,
// or nil if the logger is healthy.
func (lw *LogWriter) Status() error {
return lw.state
}
// SetLevel changes the log level.
func (lw *LogWriter) SetLevel(l Level) {
lw.lvl = l
}
// Close is a no-op that satisfies the Logger interface.
func (lw *LogWriter) Close() {}
|