couchdb-lager.git
4 years agoMove dependencies to the top level repository master
Paul J. Davis [Mon, 6 Oct 2014 22:15:17 +0000 (17:15 -0500)] 
Move dependencies to the top level repository

4 years agoremove binary rebar
Robert Newson [Wed, 27 Aug 2014 20:15:06 +0000 (21:15 +0100)] 
remove binary rebar

4 years agoRemove unnecessary version specifier import-master
Paul J. Davis [Mon, 25 Aug 2014 07:48:43 +0000 (02:48 -0500)] 
Remove unnecessary version specifier

4 years agoUse https for dependencies
Paul J. Davis [Mon, 25 Aug 2014 07:37:31 +0000 (02:37 -0500)] 
Use https for dependencies

4 years agofix path to goldrush in rebar.config
benoitc [Thu, 13 Feb 2014 16:54:03 +0000 (17:54 +0100)] 
fix path to goldrush in rebar.config

4 years agopoint to apache goldrush.
benoitc [Thu, 13 Feb 2014 16:50:53 +0000 (17:50 +0100)] 
point to apache goldrush.

4 years agoMerge pull request #201 from yfyf/flush_only_events
Andrew Thompson [Wed, 12 Feb 2014 19:07:01 +0000 (14:07 -0500)] 
Merge pull request #201 from yfyf/flush_only_events

Only discard gen_event notifications on high watermark

4 years agoOnly discard gen_event notifications on high watermark
Ignas Vyšniauskas [Wed, 12 Feb 2014 15:27:45 +0000 (16:27 +0100)] 
Only discard gen_event notifications on high watermark

Previously all messages were being flushed, which meant that trapped exits
and internal gen_event messaging was also flushed, leading to undefined
behaviour.

closes #198

4 years agoMerge pull request #200 from yfyf/fix_tests
Andrew Thompson [Wed, 12 Feb 2014 16:05:58 +0000 (11:05 -0500)] 
Merge pull request #200 from yfyf/fix_tests

Fix: correct lager startup in crash_log test

4 years agoFix: correct lager startup in crash_log test
Ignas Vyšniauskas [Wed, 12 Feb 2014 14:44:06 +0000 (15:44 +0100)] 
Fix: correct lager startup in crash_log test

4 years agoMerge pull request #199 from howleysv/pr-empty-tuple-fix
Andrew Thompson [Tue, 11 Feb 2014 16:18:41 +0000 (11:18 -0500)] 
Merge pull request #199 from howleysv/pr-empty-tuple-fix

Fixed empty tuple bug in lager:pr/2

4 years agoFixed issue whereby element(1, ...) is called on every tuple nested inside a pretty...
Shane Howley [Tue, 11 Feb 2014 12:38:48 +0000 (12:38 +0000)] 
Fixed issue whereby element(1, ...) is called on every tuple nested inside a pretty printed record regardless of size.
Modified unit test to test 0 size tuple case.

4 years agoMerge pull request #196 from basho/adt-less-proplists
Andrew Thompson [Sat, 8 Feb 2014 05:13:30 +0000 (00:13 -0500)] 
Merge pull request #196 from basho/adt-less-proplists

Don't use the proplists module when decoding error_logger messages

4 years agoAdd comment and fix EQC generator
Andrew Thompson [Fri, 7 Feb 2014 20:30:01 +0000 (15:30 -0500)] 
Add comment and fix EQC generator

4 years agoMerge pull request #197 from basho/gh-164
Andrew Thompson [Fri, 7 Feb 2014 19:50:54 +0000 (14:50 -0500)] 
Merge pull request #197 from basho/gh-164

Add newline to error_msg:error_report lines in crash.log, see #164

4 years agoAdd newline to error_msg:error_report lines in crash.log, see #164
Andrew Thompson [Fri, 7 Feb 2014 19:29:13 +0000 (14:29 -0500)] 
Add newline to error_msg:error_report lines in crash.log, see #164

4 years agoMerge pull request #170 from MaximMinin/master
Andrew Thompson [Thu, 6 Feb 2014 20:35:45 +0000 (15:35 -0500)] 
Merge pull request #170 from MaximMinin/master

pretty printing of nested records

4 years agoMerge pull request #193 from zuav/master
Andrew Thompson [Thu, 6 Feb 2014 20:27:06 +0000 (15:27 -0500)] 
Merge pull request #193 from zuav/master

Fix lager_console_backend:is_new_style_console_available() function

4 years agoDon't use the proplists module when decoding error_logger messages
Andrew Thompson [Mon, 3 Feb 2014 18:32:31 +0000 (13:32 -0500)] 
Don't use the proplists module when decoding error_logger messages

Proplist module is a lot slower than lists:keyfind, which is a BIF,
because proplists has to work with 'bare' atoms as well as 2-tuples.

This should marginally improve the throughput when printing many
error_logger messages.

4 years agoMerge pull request #194 from basho/feature/xref
Andrew Thompson [Tue, 28 Jan 2014 19:31:08 +0000 (11:31 -0800)] 
Merge pull request #194 from basho/feature/xref

Add xref target, with exclusions

4 years agoMake xref depend on compile
Andrew Thompson [Tue, 28 Jan 2014 19:31:26 +0000 (14:31 -0500)] 
Make xref depend on compile

4 years agoAdd xref target, with exclusions
Andrew Thompson [Tue, 28 Jan 2014 19:14:39 +0000 (14:14 -0500)] 
Add xref target, with exclusions

4 years agoFix lager_console_backend:is_new_style_console_available() function: check for win32...
Alexander Zhukov [Sat, 25 Jan 2014 20:32:22 +0000 (00:32 +0400)] 
Fix lager_console_backend:is_new_style_console_available() function: check for win32 was incorrect.

4 years agoRoll version 2.0.2 2.0.2
Andrew Thompson [Fri, 24 Jan 2014 05:42:41 +0000 (00:42 -0500)] 
Roll version 2.0.2

4 years agoMerge pull request #192 from basho/adt-fix-empty-binary-printing
Andrew Thompson [Fri, 24 Jan 2014 05:32:25 +0000 (21:32 -0800)] 
Merge pull request #192 from basho/adt-fix-empty-binary-printing

Remove the workaround for the bug when printing empty binaries in W mode

4 years agoForce goldrush to version 0.1.6
Andrew Thompson [Fri, 24 Jan 2014 05:39:08 +0000 (00:39 -0500)] 
Force goldrush to version 0.1.6

4 years agoFix test run on OSX
Andrew Thompson [Fri, 24 Jan 2014 05:28:27 +0000 (00:28 -0500)] 
Fix test run on OSX

4 years agoRemove the workaround for the bug when printing empty binaries in W mode
Andrew Thompson [Fri, 24 Jan 2014 05:09:26 +0000 (00:09 -0500)] 
Remove the workaround for the bug when printing empty binaries in W mode

The fix for this was merged in R16 somewhere, so this workaround is no
longer needed, and fixes the EQC test from spotting the divergent
behaviour with io_lib.

See also: http://erlang.org/pipermail/erlang-patches/2012-September/003034.html

4 years agoMerge pull request #188 from basho/adt-dialyzer-fixes
Andrew Thompson [Fri, 24 Jan 2014 05:05:13 +0000 (21:05 -0800)] 
Merge pull request #188 from basho/adt-dialyzer-fixes

Fix dialyzer warnings and make all the records typed

4 years agoFix dialyzer warnings and make all the records typed
Andrew Thompson [Tue, 14 Jan 2014 22:04:25 +0000 (17:04 -0500)] 
Fix dialyzer warnings and make all the records typed

4 years agoTweak test and update goldrush dep
Andrew Thompson [Fri, 24 Jan 2014 04:37:28 +0000 (23:37 -0500)] 
Tweak test and update goldrush dep

4 years agoAdd tracing support for metadata notfound
Pedram Nimreezi [Thu, 23 Jan 2014 23:15:39 +0000 (18:15 -0500)] 
Add tracing support for metadata notfound

5 years agoMerge pull request #187 from basho/adt-dialyzer-mad-science
Andrew Thompson [Tue, 14 Jan 2014 21:52:10 +0000 (13:52 -0800)] 
Merge pull request #187 from basho/adt-dialyzer-mad-science

Rework how dialyzer PLTs are built and used

5 years agoRefactor into tools.mk
Andrew Thompson [Tue, 14 Jan 2014 20:30:40 +0000 (15:30 -0500)] 
Refactor into tools.mk

5 years agoAdd plt to gitignore
Andrew Thompson [Tue, 14 Jan 2014 19:58:03 +0000 (14:58 -0500)] 
Add plt to gitignore

5 years agoRework how dialyzer PLTs are built and used
Andrew Thompson [Tue, 14 Jan 2014 18:11:36 +0000 (13:11 -0500)] 
Rework how dialyzer PLTs are built and used

This commit splits the PLTs into 2, one for all the required OTP
applications that are in the stdlib, and the other for the rebar
dependancies. Each one is created if it does not exist, then checked for
validity and then --add_to_plt is used to add any missing files (this is
very fast if nothing needs to be added). Then the application is
dialyzed using both PLTs.

The 'combo' PLT which resides in ~ is intended to be used by all of
Riak's deps, so it can grow to cover the set of OTP applications that
Riak depends on. Each project can only specify the ones it cares about.

5 years agoMerge pull request #186 from basho/feature/variable-plt-target
Andrew Thompson [Tue, 14 Jan 2014 02:16:52 +0000 (18:16 -0800)] 
Merge pull request #186 from basho/feature/variable-plt-target

Allow PLT destination to be specified

5 years agoUpdate PHONY Makefile target
Reid Draper [Tue, 7 Jan 2014 20:04:46 +0000 (14:04 -0600)] 
Update PHONY Makefile target

5 years agoAllow PLT destination to be specified
Reid Draper [Sat, 4 Jan 2014 00:15:59 +0000 (18:15 -0600)] 
Allow PLT destination to be specified

5 years agoMerge pull request #185 from filmor/patch-1
Andrew Thompson [Thu, 2 Jan 2014 17:37:17 +0000 (09:37 -0800)] 
Merge pull request #185 from filmor/patch-1

Fixed formatting.

5 years agoFixed formatting.
Benedikt Sauer [Thu, 2 Jan 2014 14:35:25 +0000 (15:35 +0100)] 
Fixed formatting.

5 years agoMerge pull request #179 from DeadZen/dz-fix-eunit-startup
Andrew Thompson [Mon, 11 Nov 2013 21:55:05 +0000 (13:55 -0800)] 
Merge pull request #179 from DeadZen/dz-fix-eunit-startup

Fix lager eunit initialization

5 years agoFix lager eunit initialization
Pedram Nimreezi [Mon, 11 Nov 2013 21:50:24 +0000 (16:50 -0500)] 
Fix lager eunit initialization

5 years agoRoll version 2.0.1 2.0.1
Andrew Thompson [Fri, 8 Nov 2013 20:57:19 +0000 (15:57 -0500)] 
Roll version 2.0.1

5 years agoUpdate to goldrush 0.1.5
Andrew Thompson [Fri, 8 Nov 2013 20:57:04 +0000 (15:57 -0500)] 
Update to goldrush 0.1.5

5 years agoMerge pull request #178 from josephwecker/master
Andrew Thompson [Fri, 8 Nov 2013 20:36:17 +0000 (12:36 -0800)] 
Merge pull request #178 from josephwecker/master

Set source encoding so R17 compiler doesn't choke on lager_file_backend

5 years agoSet source encoding so R17 compiler doesn't choke on code introduced in 2b544c4
Joseph Wecker [Fri, 8 Nov 2013 20:02:55 +0000 (12:02 -0800)] 
Set source encoding so R17 compiler doesn't choke on code introduced in 2b544c4

5 years agoMerge pull request #167 from essen/update-rebar
Andrew Thompson [Fri, 8 Nov 2013 19:33:13 +0000 (11:33 -0800)] 
Merge pull request #167 from essen/update-rebar

Update rebar to support REBAR_DEPS_DIR

5 years agoMerge pull request #173 from mururu/fix-doc
Sean Cribbs [Sun, 6 Oct 2013 22:40:40 +0000 (15:40 -0700)] 
Merge pull request #173 from mururu/fix-doc

Fix doc of lager_console_backend

5 years agoFix doc of lager_console_backend
Yuki Ito [Thu, 3 Oct 2013 05:34:43 +0000 (14:34 +0900)] 
Fix doc of lager_console_backend

5 years agopr fun for nested records.
Maxim Minin [Mon, 2 Sep 2013 17:13:59 +0000 (19:13 +0200)] 
pr fun for nested records.

5 years agosimple test for pretty print of nested records.
Maxim Minin [Mon, 2 Sep 2013 17:05:30 +0000 (19:05 +0200)] 
simple test for pretty print of nested records.

5 years agoUpdate rebar to support REBAR_DEPS_DIR
Loïc Hoguin [Tue, 13 Aug 2013 20:33:59 +0000 (22:33 +0200)] 
Update rebar to support REBAR_DEPS_DIR

5 years agoMerge pull request #165 from ferd/master
Andrew Thompson [Thu, 8 Aug 2013 16:02:32 +0000 (09:02 -0700)] 
Merge pull request #165 from ferd/master

Clarifying how to disable 'crash_log'

5 years agoClarifying how to disable 'crash_log'
Fred Hebert [Thu, 8 Aug 2013 15:58:53 +0000 (11:58 -0400)] 
Clarifying how to disable 'crash_log'

The current set of instructions mention 'crash_log' being left undefined
as a variable to disable it, but when that happens, lager starts it at a
default location.

This commit makes it obvious that the value must explicitly set to
'undefined' in order to disable crash_log.

5 years agoMerge pull request #156 from onlychoice/patch-1
Andrew Thompson [Thu, 27 Jun 2013 19:39:47 +0000 (12:39 -0700)] 
Merge pull request #156 from onlychoice/patch-1

Update README: fix no_file error and config format error

5 years agoAdd a test for rotation failing and returning an error
Andrew Thompson [Thu, 27 Jun 2013 19:45:02 +0000 (15:45 -0400)] 
Add a test for rotation failing and returning an error

5 years agoUpdate README: fix no_file error and config format error
Zhihui Jiao [Sat, 22 Jun 2013 17:34:11 +0000 (01:34 +0800)] 
Update README: fix no_file error and config format error

5 years agoFixed log rotation hang when file:delete fails.
Shane Howley [Thu, 20 Jun 2013 14:40:35 +0000 (15:40 +0100)] 
Fixed log rotation hang when file:delete fails.

5 years agoMerge pull request #151 from basho/adt-disable-date-rotation
Andrew Thompson [Thu, 13 Jun 2013 18:24:53 +0000 (11:24 -0700)] 
Merge pull request #151 from basho/adt-disable-date-rotation

Can't disable periodic file rotation

5 years agoMerge pull request #150 from seriyps/async-window
Andrew Thompson [Thu, 13 Jun 2013 18:19:47 +0000 (11:19 -0700)] 
Merge pull request #150 from seriyps/async-window

Add `async_threshold_window` option. #147

5 years agoMerge pull request #148 from h4cc/patch-1
Andrew Thompson [Thu, 13 Jun 2013 18:02:28 +0000 (11:02 -0700)] 
Merge pull request #148 from h4cc/patch-1

Removed incorrect ']}' from example config.

5 years agoAllow date rotation to be disabled
Andrew Thompson [Thu, 13 Jun 2013 17:41:00 +0000 (13:41 -0400)] 
Allow date rotation to be disabled

5 years agoAdd `async_threshold_window` option. #147
Sergey Prokhorov [Fri, 7 Jun 2013 18:15:08 +0000 (22:15 +0400)] 
Add `async_threshold_window` option. #147

5 years agoRemoved incorrect ']}' from example config.
h4cc [Mon, 3 Jun 2013 13:36:48 +0000 (16:36 +0300)] 
Removed incorrect ']}' from example config.

5 years agoMerge pull request #146 from dLuna/master
Andrew Thompson [Tue, 14 May 2013 20:51:47 +0000 (13:51 -0700)] 
Merge pull request #146 from dLuna/master

Documentation spelling fix.

5 years agoRemove spurious ]
Daniel Luna [Tue, 14 May 2013 17:31:24 +0000 (13:31 -0400)] 
Remove spurious ]

5 years agoUpdate goldrush, remove syntax_tools as an included application. Closes #140 2.0.0
Andrew Thompson [Mon, 6 May 2013 21:25:21 +0000 (17:25 -0400)] 
Update goldrush, remove syntax_tools as an included application. Closes #140

5 years agoMerge pull request #143 from basho/adt-win32-console
Andrew Thompson [Fri, 3 May 2013 20:42:01 +0000 (13:42 -0700)] 
Merge pull request #143 from basho/adt-win32-console

Allow windows users to use the 'old shell' as there's no alternatives

5 years agoAllow windows users to use the 'old shell' as there's no alternatives
Andrew Thompson [Fri, 3 May 2013 20:01:26 +0000 (16:01 -0400)] 
Allow windows users to use the 'old shell' as there's no alternatives

5 years agoMerge pull request #139 from basho/adt-avoid-console-io-hang
Andrew Thompson [Wed, 24 Apr 2013 02:25:06 +0000 (19:25 -0700)] 
Merge pull request #139 from basho/adt-avoid-console-io-hang

Disable console backend if oldshell is used

5 years agoRework the 'old shell' warning, and some related work
Andrew Thompson [Wed, 24 Apr 2013 01:23:36 +0000 (21:23 -0400)] 
Rework the 'old shell' warning, and some related work

Add ability for handlers to 'fatally' fail, so they won't attempt to
reinstall themselves later.

Also fix an issue with the INT_LOG macro not doing the right thing when
only the lager_throttle_backend was installed.

5 years agoDisable I/O to the console if the old-style Erlang console is used
Scott Lystig Fritchie [Tue, 23 Apr 2013 15:19:14 +0000 (10:19 -0500)] 
Disable I/O to the console if the old-style Erlang console is used

There's a nasty problem with the Erlang VM + lager when the old-style
Erlang console is used.  You can use the "-oldshell" flag to explicitly
get the old-style shell.  However, if the Erlang VM is started when
*not* associated with a pseudo-tty, the VM will silently use the
old-style shell (because the new-style shell requires a pty to
support command line editing, etc.).  The most common way of starting
the Erlang VM without a pty is to start it via a non-interactive
SSH session.

This patch is opinionated in what to do in the case when the old-
style shell is detected.  My opinion is:

1. Shout loudly to the console log (in a separate Erlang process, so
   that the main lager event handler proc won't be blocked by any
   problems with the old-style console).  It's almost certainly
   likely that you really don't want to run Erlang with the old-style
   shell.  But many sysadmins don't look closely at their systems'
   log files, so (for example) shouting the same message 10x in a row
   is legit.  Reviewers: YMMV.
2. Set a SASL alarm.  Again, many sysadmins' are bad at looking at
   log messages.  Setting a SASL alarm is another method to try to
   get their attention.

To test, create a test script called `/tmp/testit.sh`, changing the
`-pz` parameter to point to the correct place for lager's `ebin`
subdirectory:

    #!/bin/sh

    erl -pz /Users/fritchie/b/src/lager/ebin -eval '{application:start(sasl), lager_console_backend:init([error, true]), timer:sleep(5*1000), io:format("\n\nInfo: ~p\n\n", [{{alarms, alarm_handler:get_alarms()}, {user_drv, whereis(user_drv)}}]), erlang:halt()}.'

Then run it twice, using the following:

    First time: ssh -t localhost sh /tmp/testit.sh
    Second time: ssh localhost sh /tmp/testit.sh

The last lines of the first time should look like:

    Info: {{alarms,[]},{user_drv,<0.22.0>}}

The last lines of the second time should look like:

    Info: {{alarms,
               [{lager_console_backend,
                    "WARNING: old-style console is in use, so lager_console_backend log output to the console is disabled.  Restart the VM on a pseudo-tty to ensure use of the new-style VM console."}]},
           {user_drv,undefined}}

5 years agoMerge branch 'srs-persistent-traces'
Andrew Thompson [Mon, 22 Apr 2013 04:43:23 +0000 (00:43 -0400)] 
Merge branch 'srs-persistent-traces'

5 years agoAdd a simple test for persistent traces
Andrew Thompson [Mon, 22 Apr 2013 04:41:32 +0000 (00:41 -0400)] 
Add a simple test for persistent traces

5 years agoImplement persistent traces in config file
Steve Strong [Mon, 22 Apr 2013 04:39:29 +0000 (00:39 -0400)] 
Implement persistent traces in config file

5 years agoKill org-mode in favor of markdown, update docs and defaults 2.0.0rc2
Andrew Thompson [Thu, 4 Apr 2013 05:00:12 +0000 (01:00 -0400)] 
Kill org-mode in favor of markdown, update docs and defaults

5 years agoMerge branch 'DeadZen-dz-event-stream-processing'
Andrew Thompson [Fri, 19 Apr 2013 20:25:22 +0000 (16:25 -0400)] 
Merge branch 'DeadZen-dz-event-stream-processing'

5 years agoPin the goldrush dep to a SHA
Andrew Thompson [Fri, 19 Apr 2013 20:24:04 +0000 (16:24 -0400)] 
Pin the goldrush dep to a SHA

5 years agoCorrect the tracing statistics to account for # of installed traces
Andrew Thompson [Fri, 19 Apr 2013 20:20:57 +0000 (16:20 -0400)] 
Correct the tracing statistics to account for # of installed traces

5 years agoMerge branch 'dz-event-stream-processing' of git://github.com/DeadZen/lager into...
Andrew Thompson [Fri, 19 Apr 2013 19:49:54 +0000 (15:49 -0400)] 
Merge branch 'dz-event-stream-processing' of git://github.com/DeadZen/lager into DeadZen-dz-event-stream-processing

Conflicts:
src/lager_util.erl

5 years agoMerge pull request #129 from basho/adt-unicode-fix
Andrew Thompson [Mon, 15 Apr 2013 03:51:09 +0000 (20:51 -0700)] 
Merge pull request #129 from basho/adt-unicode-fix

Support writing unicode to files, thanks @jacktang

5 years agoMerge branch 'garret-smith-gs-typespec-fixes'
Andrew Thompson [Mon, 15 Apr 2013 03:25:08 +0000 (23:25 -0400)] 
Merge branch 'garret-smith-gs-typespec-fixes'

5 years agoClarify the file backend's init() spec, correct some documentation
Andrew Thompson [Mon, 15 Apr 2013 03:24:29 +0000 (23:24 -0400)] 
Clarify the file backend's init() spec, correct some documentation

5 years agoMerge pull request #130 from gomoripeti/pgö-rotate-check-interval-test
Andrew Thompson [Mon, 15 Apr 2013 02:36:44 +0000 (19:36 -0700)] 
Merge pull request #130 from gomoripeti/pgö-rotate-check-interval-test

test internal log rotation when check_interval > 0

5 years agoAdd a latin-1 test
Andrew Thompson [Fri, 12 Apr 2013 16:24:16 +0000 (12:24 -0400)] 
Add a latin-1 test

5 years agoFix typespecs for a clean dialyzer run
gsmith [Tue, 9 Apr 2013 23:46:24 +0000 (16:46 -0700)] 
Fix typespecs for a clean dialyzer run

Fix typespec of lager_file_backend:init() and lager_msg record

5 years agoImplement event stream processing
Pedram Nimreezi [Thu, 21 Mar 2013 03:19:11 +0000 (23:19 -0400)] 
Implement event stream processing

Adds transparent event stream processing and statistics.
A new 3-tuple trace is introduced as `{Key, Op, Value}`, but
for backwards compatibility `{Key, Val}` implies `=` for `Op`
and `{Key, '*'}` remains as is in the case of wildcards.
A simplified query tree module is generated which reduces
redundant selection conditions to minimize filtering overhead.

5 years agotest internal log rotation when check_interval > 0
Péter Gömöri [Fri, 5 Apr 2013 13:32:44 +0000 (15:32 +0200)] 
test internal log rotation when check_interval > 0

5 years agoMerge pull request #128 from gomoripeti/pgö-rotate-check-interval
Andrew Thompson [Thu, 4 Apr 2013 04:21:46 +0000 (21:21 -0700)] 
Merge pull request #128 from gomoripeti/pgö-rotate-check-interval

fix internal log rotation when check_interval > 0

5 years agofix internal log rotation when check_interval > 0
Péter Gömöri [Tue, 2 Apr 2013 15:55:50 +0000 (17:55 +0200)] 
fix internal log rotation when check_interval > 0

5 years agoSupport writing unicode to files, thanks @jacktang
Andrew Thompson [Wed, 3 Apr 2013 05:04:20 +0000 (01:04 -0400)] 
Support writing unicode to files, thanks @jacktang

5 years agoMerge pull request #120 from basho/adt-msg-has-datetime 2.0.0rc1
Andrew Thompson [Fri, 22 Mar 2013 15:48:42 +0000 (08:48 -0700)] 
Merge pull request #120 from basho/adt-msg-has-datetime

Improve the performance of the file backend

5 years agoTell travis to test R15B03 and R16B as well
Andrew Thompson [Tue, 19 Mar 2013 20:30:29 +0000 (16:30 -0400)] 
Tell travis to test R15B03 and R16B as well

5 years agoMerge pull request #124 from DeadZen/dz-lager-colors
Andrew Thompson [Tue, 19 Mar 2013 20:28:27 +0000 (13:28 -0700)] 
Merge pull request #124 from DeadZen/dz-lager-colors

This change implements a color spectrum for severity levels, disabled by default.

5 years agoAdd colors to lager, not enabled by default
Pedram Nimreezi [Thu, 13 Dec 2012 12:29:19 +0000 (07:29 -0500)] 
Add colors to lager, not enabled by default

5 years agoMerge pull request #121 from basho/adt-pdict-metadata
Andrew Thompson [Tue, 19 Mar 2013 18:53:54 +0000 (11:53 -0700)] 
Merge pull request #121 from basho/adt-pdict-metadata

Allow storage of persistant per-process lager metadata

5 years agoImprove the performance of the file backend
Andrew Thompson [Tue, 12 Mar 2013 20:15:59 +0000 (16:15 -0400)] 
Improve the performance of the file backend

This is done via a combination of several things:
* Make the loglevel that triggers a sync configurable
* Make the delayed_write size and intervals configurable
* Make the interval at which external rotation is checked for
  configurable
* Store the timestamp a lager_msg was created inside the lager_msg

To support these changes, several other things had to be modified:
* lager_msg:timestamp now returns a timestamp
* lager_msg:datetime was added to return the {date, time} of a message,
  like lager_msg:timestamp used to
* The configuration syntax for file backends was changed to be of the
  form {lager_file_backend, proplist()} and the old syntax was
  deprecated

Additionally, the defaults for the check_interval was raised from
'always' to 1 second, and the sync_interval was changed from 2 seconds
to one second.

5 years agoAllow storage of persistant per-process lager metadata
Andrew Thompson [Fri, 15 Mar 2013 04:48:42 +0000 (00:48 -0400)] 
Allow storage of persistant per-process lager metadata

For persistant processes with some immutable metadata (riak vnode and
the vnode ID, for example), implement lager:md/0 and lager:md/1 for
getting/setting such metadata into the process dictionary.

Such metadata is automatically included in any lager message metadata,
so you can just set it in your init() function or whatever and not have
to worry about passing the data around and using it in every lager call.

5 years agoCorrect some dialyzer warnings
Andrew Thompson [Fri, 15 Mar 2013 03:04:12 +0000 (23:04 -0400)] 
Correct some dialyzer warnings