-
Notifications
You must be signed in to change notification settings - Fork 7
/
Copy pathHistory.txt
178 lines (114 loc) · 6.07 KB
/
History.txt
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
== 3.3.13 2013-3-05
* Return any database connections used by a worker back to the pool
== 3.3.12 2013-2-24
* add nohup back to the restart_workers cap task
== 3.3.11 2013-2-21
* Use the proper methods for the routes.
== 3.3.10 2013-2-14
* Use resque_worker_cap on all UI actions.
== 3.3.9 2013-2-13
* Modified resque:work's pre-requisite :preload task to properly load a rails 4 app.
* Override the reconnect method on worker to make it thread safe.
* Added the resque_worker_rake and resque_worker_cap attributes to the config so you can override the default rake and cap behavior when starting workers from the UI.
== 3.3.8 2013-2-5
* Changed the chained system calls to use && so they all run in one call to prevent possible profile changes in a subsequent call that prevents the system from finding the cap executable.
== 3.3.7 2013-2-3
* Clean up the nohup calls. nohup only spawns another process when output is redirected 2>&1.
== 3.3.6 2013-2-3
* base.include throws and error on jruby. Use base.class_eval instead.
== 3.3.5 2013-2-3
* Use alias_method_chain on overridden worker methods so we don't have to rewrite the entire method. Prevents breaking changes when resque gets upgraded.
* Added a dependency on semantic_logger for better log messages
* Add a sleep to the work cap task so cap will respect the & with the nohup and release the process.
== 3.3.4 2013-1-30
* Add an & to the end of the resque work cap task so the nohup will release the process.
== 3.3.3 2013-12-20
* Add capistrano as a dependency
* Bug fixes
== 3.3.2 2013-11-1
* Fixed test suite so it will run under rake.
* Fixed status and statuses actions and tests
* Preload the app for workers. Added a new denpendency on the worker rake task to remove the controllers from the eager_load_paths. They aren't needed for workers and may cause load errors.
== 3.3.1 2013-10-31
* Test suite added. Fixed a few bugs around resque_scheduler, and pausing and stopping workers. Thank you very much Andrew Medeiros!
== 3.3.0 2013-08-28
* Change name to resque_manager
* Convert the gem to a true mountable rails engine
* Use upgraded resque_status gem
* Make compatible with rails 3.2
== 3.2.4 2011-12-08
* Override the filter_backtrace method for the redis backend failure object introduced in resque 1.18.0.
Filter needs to be different for resque_ui, otherwise backtrace is always an empty array on the cleaner page.
== 3.2.3 2011-11-15
* Fixes resque-cleaner UI bug. It was always showing Just now for when the job was retried.
== 3.2.2 2011-11-15
* Initialize the worker to nil.
== 3.2.1 2011-11-14
* I know that was in there before.
== 3.2.0 2011-11-14
* Added the ability to pause workers from the UI.
* If the worker is processing a job, it will pause the job on the next call to tick.
* JobWithStatus class now exposed the worker so the job can manually inspect the worker's status if necessary.
== 3.1.7 2011-11-03
* It seems that when jruby workers are started in server mode, they don't accept the -INT signal.
Changed the logic to send the -TERM signal if jruby, and -QUIT otherwise.
-TERM gracefully kills the main pid and does a -9 on the child if there is one.
Since jruby doesn't fork a child, the main worker is gracefully killed.
== 3.1.6 2011-11-02
* Need the resque:restart cap task to pass the resque_rake value to the resque:restart rake task so it can start workers with the modified rake options
== 3.1.5 2011-11-01
* Added the ability to set an optional :resque_rake variable in your deploy.rb file that is used by the resque:work task if set.
-This allows you to set alternate options, particularly useful for jruby.
ex.
go through jexec so we get the right path and java settings
set :rake, "script/jexec rake"
set :resque_rake, "script/jexec -p rake" #extra memory settings for resque workers
-the jexec file in script/ may look like:
scriptdir=`dirname $0`
# Hack to get things working under capistrano (used for torquebox to get rake working, still needed?)
if [ "$1" = "-j" ]; then
shift
export JRUBY_HOME=$1
shift
fi
if [ "$1" = "-m" ]; then
shift
export JAVA_MEM="-Xmx2048m -Xms256m -XX:PermSize=1024m -XX:MaxPermSize=1024m"
export JAVA_STACK=-Xss4096k
fi
export PATH=$scriptdir:/opt/jruby/bin:$PATH
# Configuration settings for all Rake Tasks and Resque Workers
if [ "$1" = "-p" ]; then
shift
# Set Heap Space for Young/Eden GC: 512MB
# Initial Heap size: 2GB
# Max Heap size: 4GB
# Use server JVM
# PermGenSize: 64MB
# Max PermGenSize: 128MB
# Thread Stack Size: 1024k
export JRUBY_OPTS="-J-Xmn512m -J-Xms2048m -J-Xmx4096m -J-server -J-XX:PermSize=64m -J-XX:MaxPermSize=128m -J-Xss1024k"
echo "JRUBY_OPTS=${JRUBY_OPTS}"
nohup $* & #cap task runs in background
else
exec $*
fi
== 3.1.4 2011-10-27
* Changed all the cap tasks to use the :resque role.
- You now need to define the :resque role in your deploy/<environment>.rb file with the servers that will run your workers.
role :resque, 'server1', 'server2'
The cap tasks used to run the :app role, but this has been changed to allow you to better refine your servers. You may not run
the resque workers on all, or even any of your :app servers.
== 3.1.3 2011-10-24
* Guard against nil in templates
== 3.1.2 2011-10-13
* Fix a typo
== 3.1.1 2011-10-13
* Modified the restart_workers and quit_workers cap tasks to no longer use the , :only => {:resque_restart => true} flag.
- These tasks now execute the rake tasks on all servers in the farm
* Modified the restart_workers and quit_workers rake tasks to not execute separate cap tasks for each worker.
- Task now iterates through the workers and restarts the ones on the box the rake task is executing on.
== 3.1.0 2011-09-22
* Added Resque Cleaner to manage the failed queue. Removed the Failed tab.
== 3.0.0 2011-09-16
* converted plugging to a gem/engine for Rails 3