gitlab-org--gitlab-foss/Gemfile.lock

1015 lines
25 KiB
Text
Raw Normal View History

2011-10-08 17:36:38 -04:00
GEM
remote: https://rubygems.org/
2011-10-08 17:36:38 -04:00
specs:
RedCloth (4.3.2)
ace-rails-ap (4.1.0)
2017-02-21 12:01:53 -05:00
actionmailer (4.2.8)
actionpack (= 4.2.8)
actionview (= 4.2.8)
activejob (= 4.2.8)
mail (~> 2.5, >= 2.5.4)
2015-11-25 11:18:44 -05:00
rails-dom-testing (~> 1.0, >= 1.0.5)
2017-02-21 12:01:53 -05:00
actionpack (4.2.8)
actionview (= 4.2.8)
activesupport (= 4.2.8)
2015-11-25 11:18:44 -05:00
rack (~> 1.6)
rack-test (~> 0.6.2)
2015-11-25 11:18:44 -05:00
rails-dom-testing (~> 1.0, >= 1.0.5)
rails-html-sanitizer (~> 1.0, >= 1.0.2)
2017-02-21 12:01:53 -05:00
actionview (4.2.8)
activesupport (= 4.2.8)
2014-05-29 08:13:01 -04:00
builder (~> 3.1)
erubis (~> 2.7.0)
2015-11-25 11:18:44 -05:00
rails-dom-testing (~> 1.0, >= 1.0.5)
2017-02-21 12:01:53 -05:00
rails-html-sanitizer (~> 1.0, >= 1.0.3)
activejob (4.2.8)
activesupport (= 4.2.8)
2015-11-25 11:18:44 -05:00
globalid (>= 0.3.0)
2017-02-21 12:01:53 -05:00
activemodel (4.2.8)
activesupport (= 4.2.8)
2014-05-29 08:13:01 -04:00
builder (~> 3.1)
2017-02-21 12:01:53 -05:00
activerecord (4.2.8)
activemodel (= 4.2.8)
activesupport (= 4.2.8)
2015-11-25 11:18:44 -05:00
arel (~> 6.0)
activerecord_sane_schema_dumper (0.2)
rails (>= 4, < 5)
2017-02-21 12:01:53 -05:00
activesupport (4.2.8)
2015-11-25 11:18:44 -05:00
i18n (~> 0.7)
2014-05-29 08:13:01 -04:00
minitest (~> 5.1)
2015-11-25 11:18:44 -05:00
thread_safe (~> 0.3, >= 0.3.4)
2014-05-29 08:13:01 -04:00
tzinfo (~> 1.1)
acts-as-taggable-on (4.0.0)
activerecord (>= 4.0)
addressable (2.3.8)
2015-11-25 11:18:44 -05:00
after_commit_queue (1.3.0)
activerecord (>= 3.0)
akismet (2.0.0)
allocations (1.0.5)
2017-02-21 12:01:53 -05:00
arel (6.0.4)
2015-11-25 17:03:30 -05:00
asana (0.4.0)
faraday (~> 0.9)
faraday_middleware (~> 0.9)
faraday_middleware-multi_json (~> 0.0)
oauth2 (~> 1.0)
2015-11-25 11:18:44 -05:00
asciidoctor (1.5.3)
asciidoctor-plantuml (0.0.7)
asciidoctor (~> 1.5)
ast (2.3.0)
attr_encrypted (3.0.3)
encryptor (~> 3.0.0)
2014-12-19 09:15:29 -05:00
attr_required (1.0.0)
autoparse (0.3.3)
addressable (>= 2.3.1)
extlib (>= 0.9.15)
multi_json (>= 1.0.0)
2016-01-01 21:11:39 -05:00
autoprefixer-rails (6.2.3)
2015-02-19 18:02:49 -05:00
execjs
json
2013-09-29 08:44:49 -04:00
awesome_print (1.2.0)
2015-08-25 21:42:46 -04:00
axiom-types (0.1.1)
descendants_tracker (~> 0.0.4)
ice_nine (~> 0.11.0)
thread_safe (~> 0.3, >= 0.3.1)
babosa (1.0.2)
2016-05-02 07:29:17 -04:00
base32 (0.3.2)
bcrypt (3.1.11)
benchmark-ips (2.3.0)
2013-09-29 08:44:49 -04:00
better_errors (1.0.1)
coderay (>= 1.0.0)
2013-05-01 06:29:29 -04:00
erubis (>= 2.6.6)
2013-06-24 15:03:32 -04:00
binding_of_caller (0.7.2)
debug_inspector (>= 0.0.1)
2016-01-01 21:11:39 -05:00
bootstrap-sass (3.3.6)
autoprefixer-rails (>= 5.2.1)
sass (>= 3.3.4)
2017-01-18 12:21:57 -05:00
brakeman (3.4.1)
browser (2.2.0)
2017-02-21 12:01:53 -05:00
builder (3.2.3)
bullet (5.2.0)
activesupport (>= 3.0.0)
uniform_notifier (~> 1.10.0)
bundler-audit (0.5.0)
bundler (~> 1.2)
thor (~> 0.18)
byebug (9.0.6)
2016-03-14 01:52:19 -04:00
capybara (2.6.2)
addressable
2011-10-08 17:36:38 -04:00
mime-types (>= 1.16)
nokogiri (>= 1.3.3)
rack (>= 1.0.0)
rack-test (>= 0.5.4)
xpath (~> 2.0)
2015-08-25 21:42:46 -04:00
capybara-screenshot (1.0.11)
2015-04-25 14:10:09 -04:00
capybara (>= 1.0, < 3)
launchy
2016-03-03 20:32:18 -05:00
carrierwave (0.10.0)
activemodel (>= 3.2.0)
activesupport (>= 3.2.0)
json (>= 1.7)
2016-03-03 20:32:18 -05:00
mime-types (>= 1.16)
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
cause (0.1)
2015-11-10 10:56:05 -05:00
charlock_holmes (0.7.3)
chronic (0.10.2)
2016-05-18 16:21:51 -04:00
chronic_duration (0.10.6)
numerizer (~> 0.1.1)
2015-11-25 11:18:44 -05:00
chunky_png (1.3.5)
cliver (0.3.2)
coderay (1.1.0)
coercible (1.0.0)
descendants_tracker (~> 0.0.1)
coffee-rails (4.1.1)
2011-10-08 17:36:38 -04:00
coffee-script (>= 2.2.0)
railties (>= 4.0.0, < 5.1.x)
2015-05-29 00:05:14 -04:00
coffee-script (2.4.1)
2011-10-08 17:36:38 -04:00
coffee-script-source
execjs
coffee-script-source (1.10.0)
colorize (0.7.7)
2017-02-21 12:01:53 -05:00
concurrent-ruby (1.0.4)
connection_pool (2.2.1)
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
crack (0.4.3)
safe_yaml (~> 1.0.0)
2015-11-24 15:42:42 -05:00
creole (0.5.0)
css_parser (1.4.1)
addressable
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
d3_rails (3.5.11)
railties (>= 3.1.0)
2015-08-25 21:42:46 -04:00
daemons (1.2.3)
database_cleaner (1.5.3)
debug_inspector (0.0.2)
debugger-ruby_core_source (1.3.8)
deckar01-task_list (1.0.6)
activesupport (~> 4.0)
html-pipeline
rack (~> 1.0)
default_value_for (3.0.2)
activerecord (>= 3.2.0, < 5.1)
2015-08-25 21:42:46 -04:00
descendants_tracker (0.0.4)
thread_safe (~> 0.3, >= 0.3.1)
devise (4.2.0)
2014-07-09 07:17:45 -04:00
bcrypt (~> 3.0)
2012-07-06 02:50:24 -04:00
orm_adapter (~> 0.1)
railties (>= 4.1.0, < 5.1)
2015-09-19 21:12:32 -04:00
responders
warden (~> 1.2.3)
devise-two-factor (3.0.0)
2015-03-27 18:35:26 -04:00
activesupport
attr_encrypted (>= 1.3, < 4, != 2)
devise (~> 4.0)
2015-08-25 21:42:46 -04:00
railties
rotp (~> 2.0)
diff-lcs (1.2.5)
diffy (3.1.0)
2014-09-02 16:13:14 -04:00
docile (1.1.5)
domain_name (0.5.20161021)
unf (>= 0.0.5, < 1.0.0)
doorkeeper (4.2.0)
railties (>= 4.2)
2015-11-25 11:18:44 -05:00
dropzonejs-rails (0.7.2)
rails (> 3.1)
email_reply_trimmer (0.1.6)
email_spec (1.6.0)
2012-11-18 15:51:49 -05:00
launchy (~> 2.1)
mail (~> 2.2)
encryptor (3.0.0)
2015-08-25 21:42:46 -04:00
equalizer (0.0.11)
2011-10-08 17:36:38 -04:00
erubis (2.7.0)
escape_utils (1.1.1)
2015-08-25 21:42:46 -04:00
eventmachine (1.0.8)
excon (0.52.0)
2015-08-25 21:42:46 -04:00
execjs (2.6.0)
2014-07-28 05:47:27 -04:00
expression_parser (0.9.0)
extlib (0.9.16)
2016-12-08 01:30:07 -05:00
factory_girl (4.7.0)
2012-08-28 01:28:09 -04:00
activesupport (>= 3.0.0)
2016-12-08 01:30:07 -05:00
factory_girl_rails (4.7.0)
factory_girl (~> 4.7.0)
2012-08-28 01:28:09 -04:00
railties (>= 3.0.0)
2015-10-07 21:54:15 -04:00
faraday (0.9.2)
multipart-post (>= 1.2, < 3)
2015-08-25 21:42:46 -04:00
faraday_middleware (0.10.0)
faraday (>= 0.7.4, < 0.10)
2015-11-25 17:03:30 -05:00
faraday_middleware-multi_json (0.0.6)
faraday_middleware
multi_json
ffaker (2.4.0)
2015-08-25 21:42:46 -04:00
ffi (1.9.10)
flay (2.6.1)
ruby_parser (~> 3.0)
sexp_processor (~> 4.0)
2015-11-25 11:18:44 -05:00
flowdock (0.7.1)
httparty (~> 0.7)
multi_json
fog-aws (0.11.0)
fog-core (~> 1.38)
2016-01-11 11:41:11 -05:00
fog-json (~> 1.0)
fog-xml (~> 0.1)
ipaddress (~> 0.8)
fog-core (1.42.0)
2013-07-08 02:47:31 -04:00
builder
excon (~> 0.49)
formatador (~> 0.2)
fog-google (0.5.0)
fog-core
fog-json
fog-xml
fog-json (1.0.2)
fog-core (~> 1.0)
multi_json (~> 1.10)
fog-local (0.3.0)
2016-01-11 11:41:11 -05:00
fog-core (~> 1.27)
fog-openstack (0.1.6)
fog-core (>= 1.39)
fog-json (>= 1.0)
ipaddress (>= 0.8)
fog-rackspace (0.1.1)
fog-core (>= 1.35)
fog-json (>= 1.0)
fog-xml (>= 0.1)
ipaddress (>= 0.8)
fog-xml (0.1.2)
fog-core
nokogiri (~> 1.5, >= 1.5.11)
2016-06-13 08:05:08 -04:00
font-awesome-rails (4.6.1.0)
railties (>= 3.2, < 5.1)
2015-08-25 21:42:46 -04:00
foreman (0.78.0)
thor (~> 0.19.1)
formatador (0.2.5)
2015-06-25 21:43:24 -04:00
fuubar (2.0.0)
rspec (~> 3.0)
ruby-progressbar (~> 1.4)
gemnasium-gitlab-service (0.2.6)
2015-02-25 10:14:10 -05:00
rugged (~> 0.21)
gemojione (3.0.1)
2015-03-11 19:05:01 -04:00
json
2015-08-25 21:42:46 -04:00
get_process_mem (0.2.0)
gherkin-ruby (0.3.2)
github-linguist (4.7.6)
charlock_holmes (~> 0.7.3)
escape_utils (~> 1.1.0)
mime-types (>= 1.19)
rugged (>= 0.23.0b)
github-markup (1.4.0)
gitlab-flowdock-git-hook (1.0.1)
flowdock (~> 0.7)
gitlab-grit (>= 2.4.1)
multi_json
gitlab-grit (2.8.1)
charlock_holmes (~> 0.6)
diff-lcs (~> 1.1)
mime-types (>= 1.16, < 3)
posix-spawn (~> 0.3)
gitlab-markup (1.5.1)
2015-03-17 12:15:39 -04:00
gitlab_omniauth-ldap (1.2.1)
net-ldap (~> 0.9)
omniauth (~> 1.0)
pyu-ruby-sasl (~> 0.0.3.1)
rubyntlm (~> 0.3)
globalid (0.3.7)
2015-11-25 11:18:44 -05:00
activesupport (>= 4.1.0)
gollum-grit_adapter (1.0.1)
2015-03-20 09:15:56 -04:00
gitlab-grit (~> 2.7, >= 2.7.1)
gollum-lib (4.2.1)
github-markup (~> 1.4.0)
2015-08-25 21:42:46 -04:00
gollum-grit_adapter (~> 1.0)
nokogiri (~> 1.6.4)
rouge (~> 2.0)
sanitize (~> 2.1.0)
stringex (~> 2.5.1)
2016-02-28 07:11:43 -05:00
gollum-rugged_adapter (0.4.2)
mime-types (>= 1.15)
rugged (~> 0.24.0, >= 0.21.3)
gon (6.1.0)
2015-11-24 15:36:36 -05:00
actionpack (>= 3.0)
json
2015-11-24 15:36:36 -05:00
multi_json
request_store (>= 1.0)
google-api-client (0.8.7)
activesupport (>= 3.2, < 5.0)
addressable (~> 2.3)
autoparse (~> 0.3)
extlib (~> 0.9)
faraday (~> 0.9)
googleauth (~> 0.3)
launchy (~> 2.4)
multi_json (~> 1.10)
retriable (~> 1.4)
signet (~> 0.6)
googleauth (0.5.1)
faraday (~> 0.9)
jwt (~> 1.4)
logging (~> 2.0)
memoist (~> 0.12)
multi_json (~> 1.11)
os (~> 0.9)
signet (~> 0.7)
2016-12-13 04:48:47 -05:00
grape (0.18.0)
2012-11-18 15:51:49 -05:00
activesupport
builder
hashie (>= 2.1.0)
2012-11-18 15:51:49 -05:00
multi_json (>= 1.3.2)
multi_xml (>= 0.5.2)
2016-12-13 04:48:47 -05:00
mustermann-grape (~> 0.4.0)
2013-05-01 06:29:29 -04:00
rack (>= 1.3.0)
2012-11-18 15:51:49 -05:00
rack-accept
2013-12-09 14:50:36 -05:00
virtus (>= 1.0.0)
2016-11-21 03:09:19 -05:00
grape-entity (0.6.0)
activesupport
multi_json (>= 1.3.2)
haml (4.0.7)
tilt
haml_lint (0.18.2)
haml (~> 4.0)
rake (>= 10, < 12)
rubocop (>= 0.36.0)
sysexits (~> 1.1)
hamlit (2.6.1)
temple (~> 0.7.6)
thor
2013-03-01 08:09:11 -05:00
tilt
2016-09-13 07:09:04 -04:00
hashie (3.4.4)
health_check (2.2.1)
rails (>= 4.0)
2015-08-25 21:42:46 -04:00
hipchat (1.5.2)
2013-06-24 15:03:32 -04:00
httparty
2015-03-30 18:53:24 -04:00
mimemagic
html-pipeline (1.11.0)
activesupport (>= 2)
nokogiri (~> 1.4)
html2text (0.2.0)
nokogiri (~> 1.6)
htmlentities (4.3.4)
http (0.9.8)
addressable (~> 2.3)
http-cookie (~> 1.0)
http-form_data (~> 1.0.1)
http_parser.rb (~> 0.6.0)
http-cookie (1.0.3)
domain_name (~> 0.5)
http-form_data (1.0.1)
http_parser.rb (0.6.0)
2015-11-25 11:18:44 -05:00
httparty (0.13.7)
json (~> 1.8)
2013-02-28 14:11:12 -05:00
multi_xml (>= 0.5.2)
httpclient (2.8.2)
2017-02-21 12:01:53 -05:00
i18n (0.8.0)
2015-08-25 21:42:46 -04:00
ice_nine (0.11.1)
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
influxdb (0.2.3)
cause
json
ipaddress (0.8.3)
2016-10-19 17:58:48 -04:00
jira-ruby (1.1.2)
activesupport
oauth (~> 0.5, >= 0.5.0)
jquery-atwho-rails (1.3.2)
jquery-rails (4.1.1)
rails-dom-testing (>= 1, < 3)
railties (>= 4.2.0)
thor (>= 0.14, < 2.0)
jquery-ui-rails (5.0.5)
railties (>= 3.2.16)
2017-02-21 12:01:53 -05:00
json (1.8.6)
2016-08-01 13:38:46 -04:00
json-schema (2.6.2)
addressable (~> 2.3.8)
jwt (1.5.6)
kaminari (0.17.0)
2012-02-11 13:34:25 -05:00
actionpack (>= 3.0.0)
activesupport (>= 3.0.0)
2015-11-25 11:18:44 -05:00
kgio (2.10.0)
2016-06-03 11:15:00 -04:00
knapsack (1.11.0)
2016-05-21 21:17:15 -04:00
rake
timecop (>= 0.1.0)
kubeclient (2.2.0)
http (= 0.9.8)
recursive-open-struct (= 1.0.0)
rest-client
launchy (2.4.3)
2012-11-18 15:51:49 -05:00
addressable (~> 2.3)
letter_opener (1.4.1)
2013-06-24 15:03:32 -04:00
launchy (~> 2.2)
letter_opener_web (1.3.0)
actionmailer (>= 3.2)
letter_opener (~> 1.0)
railties (>= 3.2)
license_finder (2.1.0)
bundler
httparty
rubyzip
thor
xml-simple
licensee (8.7.0)
rugged (~> 0.24)
little-plugger (1.1.4)
logging (2.1.0)
little-plugger (~> 1.1)
multi_json (~> 1.10)
2015-11-25 11:18:44 -05:00
loofah (2.0.3)
nokogiri (>= 1.5.9)
2016-03-28 15:37:53 -04:00
mail (2.6.4)
mime-types (>= 1.16, < 4)
mail_room (0.9.1)
memoist (0.15.0)
method_source (0.8.2)
2016-09-09 09:19:48 -04:00
mime-types (2.99.3)
2015-03-30 18:53:24 -04:00
mimemagic (0.3.0)
mini_portile2 (2.1.0)
2015-08-25 21:42:46 -04:00
minitest (5.7.0)
2014-08-21 04:14:31 -04:00
mousetrap-rails (1.4.6)
multi_json (1.12.1)
multi_xml (0.5.5)
2015-10-07 21:54:15 -04:00
multipart-post (2.0.0)
2016-12-13 04:48:47 -05:00
mustermann (0.4.0)
tool (~> 0.2)
mustermann-grape (0.4.0)
mustermann (= 0.4.0)
2015-08-25 21:42:46 -04:00
mysql2 (0.3.20)
2015-11-25 11:18:44 -05:00
net-ldap (0.12.1)
net-ssh (3.0.1)
netrc (0.11.0)
newrelic_rpm (3.16.0.318)
2017-02-21 12:01:53 -05:00
nokogiri (1.6.8.1)
mini_portile2 (~> 2.1.0)
2016-05-18 16:21:51 -04:00
numerizer (0.1.1)
oauth (0.5.1)
oauth2 (1.2.0)
2015-08-25 21:42:46 -04:00
faraday (>= 0.8, < 0.10)
jwt (~> 1.0)
multi_json (~> 1.3)
multi_xml (~> 0.5)
rack (>= 1.2, < 3)
octokit (4.6.2)
sawyer (~> 0.8.0, >= 0.5.3)
oj (2.17.4)
omniauth (1.3.2)
hashie (>= 1.2, < 4)
rack (>= 1.0, < 3)
omniauth-auth0 (1.4.1)
omniauth-oauth2 (~> 1.1)
omniauth-authentiq (0.3.0)
omniauth-oauth2 (~> 1.3, >= 1.3.1)
2016-01-07 12:27:01 -05:00
omniauth-azure-oauth2 (0.0.6)
jwt (~> 1.0)
omniauth (~> 1.0)
omniauth-oauth2 (~> 1.1)
2015-11-11 23:25:31 -05:00
omniauth-cas3 (1.1.3)
addressable (~> 2.3)
nokogiri (~> 1.6.6)
omniauth (~> 1.2)
2016-09-13 07:09:04 -04:00
omniauth-facebook (4.0.0)
2015-11-03 11:58:12 -05:00
omniauth-oauth2 (~> 1.2)
2015-08-25 21:42:46 -04:00
omniauth-github (1.1.2)
2012-09-12 00:48:22 -04:00
omniauth (~> 1.0)
omniauth-oauth2 (~> 1.1)
omniauth-gitlab (1.0.2)
2015-01-27 18:37:19 -05:00
omniauth (~> 1.0)
omniauth-oauth2 (~> 1.0)
2016-06-23 10:15:10 -04:00
omniauth-google-oauth2 (0.4.1)
2015-11-25 11:18:44 -05:00
addressable (~> 2.3)
jwt (~> 1.0)
multi_json (~> 1.3)
omniauth (>= 1.1.1)
omniauth-oauth2 (~> 1.3.1)
2015-10-06 21:48:19 -04:00
omniauth-kerberos (0.3.0)
2014-12-16 06:57:40 -05:00
omniauth-multipassword
timfel-krb5-auth (~> 0.8)
2015-08-25 21:42:46 -04:00
omniauth-multipassword (0.4.2)
2014-12-16 06:57:40 -05:00
omniauth (~> 1.0)
2015-08-25 21:42:46 -04:00
omniauth-oauth (1.1.0)
2012-09-12 00:48:22 -04:00
oauth
omniauth (~> 1.0)
2015-08-25 21:42:46 -04:00
omniauth-oauth2 (1.3.1)
oauth2 (~> 1.0)
omniauth (~> 1.2)
omniauth-oauth2-generic (0.2.2)
omniauth-oauth2 (~> 1.0)
omniauth-saml (1.7.0)
omniauth (~> 1.3)
ruby-saml (~> 1.4)
2015-10-06 22:03:42 -04:00
omniauth-shibboleth (1.2.1)
omniauth (>= 1.0.0)
2015-10-06 21:42:32 -04:00
omniauth-twitter (1.2.1)
json (~> 1.3)
omniauth-oauth (~> 1.1)
2015-08-31 06:59:52 -04:00
omniauth_crowd (2.2.3)
activesupport
nokogiri (>= 1.4.4)
omniauth (~> 1.0)
2014-12-29 01:22:56 -05:00
org-ruby (0.9.12)
2014-08-13 09:45:48 -04:00
rubypants (~> 0.2)
orm_adapter (0.5.0)
os (0.9.6)
paranoia (2.2.0)
activerecord (>= 4.0, < 5.1)
parser (2.3.1.4)
ast (~> 2.2)
2015-11-25 11:18:44 -05:00
pg (0.18.4)
2016-03-07 15:03:55 -05:00
poltergeist (1.9.0)
capybara (~> 2.1)
cliver (~> 0.3.1)
multi_json (~> 1.0)
websocket-driver (>= 0.2.0)
2015-08-25 21:42:46 -04:00
posix-spawn (0.3.11)
2015-12-14 14:18:32 -05:00
powerpack (0.1.1)
premailer (1.8.6)
css_parser (>= 1.3.6)
htmlentities (>= 4.0.0)
premailer-rails (1.9.2)
actionmailer (>= 3, < 6)
premailer (~> 1.7, >= 1.7.9)
2015-11-25 11:18:44 -05:00
pry (0.10.3)
2015-08-25 21:42:46 -04:00
coderay (~> 1.1.0)
method_source (~> 0.8.1)
2013-03-27 16:21:32 -04:00
slop (~> 3.4)
pry-byebug (3.4.1)
byebug (~> 9.0)
pry (~> 0.10)
2015-08-25 21:42:46 -04:00
pry-rails (0.3.4)
pry (>= 0.9.10)
2012-01-21 13:36:14 -05:00
pyu-ruby-sasl (0.0.3.3)
2016-12-13 04:48:47 -05:00
rack (1.6.5)
2012-11-18 15:51:49 -05:00
rack-accept (0.4.5)
rack (>= 0.4)
rack-attack (4.4.1)
2013-09-24 14:13:25 -04:00
rack
2015-10-07 22:08:30 -04:00
rack-cors (0.4.0)
rack-oauth2 (1.2.3)
2014-12-19 09:15:29 -05:00
activesupport (>= 2.3)
attr_required (>= 0.0.5)
2015-08-25 21:42:46 -04:00
httpclient (>= 2.4)
2014-12-19 09:15:29 -05:00
multi_json (>= 1.3.6)
rack (>= 1.1)
2015-08-25 21:42:46 -04:00
rack-protection (1.5.3)
2011-10-08 17:36:38 -04:00
rack
rack-proxy (0.6.0)
rack
rack-test (0.6.3)
2011-10-08 17:36:38 -04:00
rack (>= 1.0)
2017-02-21 12:01:53 -05:00
rails (4.2.8)
actionmailer (= 4.2.8)
actionpack (= 4.2.8)
actionview (= 4.2.8)
activejob (= 4.2.8)
activemodel (= 4.2.8)
activerecord (= 4.2.8)
activesupport (= 4.2.8)
bundler (>= 1.3.0, < 2.0)
2017-02-21 12:01:53 -05:00
railties (= 4.2.8)
2015-11-25 11:18:44 -05:00
sprockets-rails
rails-deprecated_sanitizer (1.0.3)
activesupport (>= 4.2.0.alpha)
2017-02-21 12:01:53 -05:00
rails-dom-testing (1.0.8)
2015-11-25 11:18:44 -05:00
activesupport (>= 4.2.0.beta, < 5.0)
2017-02-21 12:01:53 -05:00
nokogiri (~> 1.6)
2015-11-25 11:18:44 -05:00
rails-deprecated_sanitizer (>= 1.0.1)
rails-html-sanitizer (1.0.3)
2015-11-25 11:18:44 -05:00
loofah (~> 2.0)
2017-02-21 12:01:53 -05:00
railties (4.2.8)
actionpack (= 4.2.8)
activesupport (= 4.2.8)
2011-10-08 17:36:38 -04:00
rake (>= 0.8.7)
thor (>= 0.18.1, < 2.0)
rainbow (2.1.0)
raindrops (0.17.0)
2016-01-18 16:39:31 -05:00
rake (10.5.0)
rblineprof (0.3.6)
debugger-ruby_core_source (~> 1.3)
rdoc (4.2.2)
json (~> 1.4)
recaptcha (3.0.0)
json
recursive-open-struct (1.0.0)
redcarpet (3.3.3)
redis (3.2.2)
redis-actionpack (5.0.1)
actionpack (>= 4.0, < 6)
redis-rack (>= 1, < 3)
redis-store (>= 1.1.0, < 1.4.0)
redis-activesupport (5.0.1)
activesupport (>= 3, < 6)
redis-store (~> 1.2.0)
2015-08-25 21:42:46 -04:00
redis-namespace (1.5.2)
2014-11-28 13:06:21 -05:00
redis (~> 3.0, >= 3.0.4)
redis-rack (1.6.0)
rack (~> 1.5)
redis-store (~> 1.2.0)
redis-rails (5.0.1)
redis-actionpack (~> 5.0.0)
redis-activesupport (~> 5.0.0)
redis-store (~> 1.2.0)
redis-store (1.2.0)
2013-08-29 13:32:40 -04:00
redis (>= 2.2)
request_store (1.3.1)
responders (2.3.0)
2016-01-18 16:39:31 -05:00
railties (>= 4.2.0, < 5.1)
rest-client (2.0.0)
http-cookie (>= 1.0.2, < 2.0)
mime-types (>= 1.16, < 4.0)
netrc (~> 0.8)
retriable (1.4.1)
2016-07-04 06:25:09 -04:00
rinku (2.0.0)
rotp (2.1.2)
rouge (2.0.7)
2015-08-25 21:42:46 -04:00
rqrcode (0.7.0)
chunky_png
2015-03-27 18:35:26 -04:00
rqrcode-rails3 (0.1.7)
rqrcode (>= 0.4.2)
rspec (3.5.0)
rspec-core (~> 3.5.0)
rspec-expectations (~> 3.5.0)
rspec-mocks (~> 3.5.0)
rspec-core (3.5.0)
rspec-support (~> 3.5.0)
rspec-expectations (3.5.0)
diff-lcs (>= 1.2.0, < 2.0)
rspec-support (~> 3.5.0)
rspec-mocks (3.5.0)
2015-06-17 18:05:48 -04:00
diff-lcs (>= 1.2.0, < 2.0)
rspec-support (~> 3.5.0)
rspec-rails (3.5.0)
actionpack (>= 3.0)
activesupport (>= 3.0)
railties (>= 3.0)
rspec-core (~> 3.5.0)
rspec-expectations (~> 3.5.0)
rspec-mocks (~> 3.5.0)
rspec-support (~> 3.5.0)
2015-11-12 04:52:20 -05:00
rspec-retry (0.4.5)
rspec-core
rspec-support (3.5.0)
rspec_profiling (0.0.5)
activerecord
pg
rails
sqlite3
rubocop (0.46.0)
parser (>= 2.3.1.1, < 3.0)
2015-12-14 14:18:32 -05:00
powerpack (~> 0.1)
rainbow (>= 1.99.1, < 3.0)
2015-12-14 14:18:32 -05:00
ruby-progressbar (~> 1.7)
unicode-display_width (~> 1.0, >= 1.0.1)
rubocop-rspec (1.9.1)
rubocop (>= 0.42.0)
2015-09-15 16:09:32 -04:00
ruby-fogbugz (0.2.1)
crack (~> 0.4)
ruby-prof (0.16.2)
2016-05-23 14:03:04 -04:00
ruby-progressbar (1.8.1)
ruby-saml (1.4.1)
nokogiri (>= 1.5.10)
ruby_parser (3.8.2)
2015-03-02 20:28:47 -05:00
sexp_processor (~> 4.1)
2015-08-25 21:42:46 -04:00
rubyntlm (0.5.2)
2014-05-28 06:23:03 -04:00
rubypants (0.2.0)
rubyzip (1.2.0)
rufus-scheduler (3.1.10)
2016-02-28 07:11:43 -05:00
rugged (0.24.0)
safe_yaml (1.0.4)
sanitize (2.1.0)
nokogiri (>= 1.4.4)
sass (3.4.22)
sass-rails (5.0.6)
railties (>= 4.0.0, < 6)
sass (~> 3.1)
sprockets (>= 2.8, < 4.0)
sprockets-rails (>= 2.0, < 4.0)
tilt (>= 1.1, < 3)
sawyer (0.8.1)
addressable (>= 2.3.5, < 2.6)
faraday (~> 0.8, < 1.0)
scss_lint (0.47.1)
rake (>= 0.9, < 11)
sass (~> 3.4.15)
seed-fu (2.3.6)
activerecord (>= 3.1)
activesupport (>= 3.1)
2015-06-24 17:13:21 -04:00
select2-rails (3.5.9.3)
2013-03-13 15:36:26 -04:00
thor (~> 0.14)
sentry-raven (2.0.2)
faraday (>= 0.7.6, < 0.10.x)
settingslogic (2.0.9)
sexp_processor (4.7.0)
2015-08-04 18:21:12 -04:00
sham_rack (1.3.6)
rack
shoulda-matchers (2.8.0)
2012-06-08 06:28:19 -04:00
activesupport (>= 3.0.0)
sidekiq (4.2.7)
2015-12-10 12:45:36 -05:00
concurrent-ruby (~> 1.0)
connection_pool (~> 2.2, >= 2.2.0)
rack-protection (>= 1.5.0)
redis (~> 3.2, >= 3.2.1)
sidekiq-cron (0.4.4)
2015-12-10 12:45:36 -05:00
redis-namespace (>= 1.5.2)
rufus-scheduler (>= 2.0.24)
sidekiq (>= 4.2.1)
sidekiq-limit_fetch (3.4.0)
sidekiq (>= 4)
signet (0.7.3)
addressable (~> 2.3)
faraday (~> 0.9)
jwt (~> 1.5)
multi_json (~> 1.10)
2016-07-04 09:02:27 -04:00
simplecov (0.12.0)
docile (~> 1.1.0)
2016-07-04 09:02:27 -04:00
json (>= 1.8, < 3)
simplecov-html (~> 0.10.0)
simplecov-html (0.10.0)
2016-11-30 09:51:48 -05:00
slack-notifier (1.5.1)
slop (3.6.0)
2015-08-25 21:42:46 -04:00
spinach (0.8.10)
colorize
gherkin-ruby (>= 0.3.2)
json
2013-05-01 06:29:29 -04:00
spinach-rails (0.2.1)
capybara (>= 2.0.0)
2012-09-10 03:42:36 -04:00
railties (>= 3)
spinach (>= 0.4)
2016-03-09 08:12:08 -05:00
spinach-rerun-reporter (0.0.2)
spinach (~> 0.8)
spring (1.7.2)
spring-commands-rspec (1.0.4)
2014-02-15 14:46:15 -05:00
spring (>= 0.9.1)
spring-commands-spinach (1.1.0)
2014-02-15 14:46:15 -05:00
spring (>= 0.9.1)
2017-02-21 12:01:53 -05:00
sprockets (3.7.1)
concurrent-ruby (~> 1.0)
rack (> 1, < 3)
2017-02-21 12:01:53 -05:00
sprockets-rails (3.2.0)
2016-03-28 15:37:53 -04:00
actionpack (>= 4.0)
activesupport (>= 4.0)
sprockets (>= 3.0.0)
sqlite3 (1.3.13)
stackprof (0.2.10)
2015-11-09 09:11:42 -05:00
state_machines (0.4.0)
state_machines-activemodel (0.4.0)
activemodel (>= 4.1, < 5.1)
2015-11-09 09:11:42 -05:00
state_machines (>= 0.4.0)
state_machines-activerecord (0.4.0)
activerecord (>= 4.1, < 5.1)
2015-11-09 09:11:42 -05:00
state_machines-activemodel (>= 0.3.0)
stringex (2.5.2)
sys-filesystem (1.1.6)
ffi
sysexits (1.2.0)
temple (0.7.7)
Fix race conditions for AuthorizedProjectsWorker There were two cases that could be problematic: 1. Because sometimes AuthorizedProjectsWorker would be scheduled in a transaction it was possible for a job to run/complete before a COMMIT; resulting in it either producing an error, or producing no new data. 2. When scheduling jobs the code would not wait until completion. This could lead to a user creating a project and then immediately trying to push to it. Usually this will work fine, but given enough load it might take a few seconds before a user has access. The first one is problematic, the second one is mostly just annoying (but annoying enough to warrant a solution). This commit changes two things to deal with this: 1. Sidekiq scheduling now takes places after a COMMIT, this is ensured by scheduling using Rails' after_commit hook instead of doing so in an arbitrary method. 2. When scheduling jobs the calling thread now waits for all jobs to complete. Solution 2 requires tracking of job completions. Sidekiq provides a way to find a job by its ID, but this involves scanning over the entire queue; something that is very in-efficient for large queues. As such a more efficient solution is necessary. There are two main Gems that can do this in a more efficient manner: * sidekiq-status * sidekiq_status No, this is not a joke. Both Gems do a similar thing (but slightly different), and the only difference in their name is a dash vs an underscore. Both Gems however provide far more than just checking if a job has been completed, and both have their problems. sidekiq-status does not appear to be actively maintained, with the last release being in 2015. It also has some issues during testing as API calls are not stubbed in any way. sidekiq_status on the other hand does not appear to be very popular, and introduces a similar amount of code. Because of this I opted to write a simple home grown solution. After all, all we need is storing a job ID somewhere so we can efficiently look it up; we don't need extra web UIs (as provided by sidekiq-status) or complex APIs to update progress, etc. This is where Gitlab::SidekiqStatus comes in handy. This namespace contains some code used for tracking, removing, and looking up job IDs; all without having to scan over an entire queue. Data is removed explicitly, but also expires automatically just in case. Using this API we can now schedule jobs in a fork-join like manner: we schedule the jobs in Sidekiq, process them in parallel, then wait for completion. By using Sidekiq we can leverage all the benefits such as being able to scale across multiple cores and hosts, retrying failed jobs, etc. The one downside is that we need to make sure we can deal with unexpected increases in job processing timings. To deal with this the class Gitlab::JobWaiter (used for waiting for jobs to complete) will only wait a number of seconds (30 by default). Once this timeout is reached it will simply return. For GitLab.com almost all AuthorizedProjectWorker jobs complete in seconds, only very rarely do we spike to job timings of around a minute. These in turn seem to be the result of external factors (e.g. deploys), in which case a user is most likely not able to use the system anyway. In short, this new solution should ensure that jobs are processed properly and that in almost all cases a user has access to their resources whenever they need to have access.
2017-01-22 12:22:02 -05:00
test_after_commit (1.1.0)
2015-08-25 21:42:46 -04:00
activerecord (>= 3.2)
thin (1.7.0)
2015-08-25 21:42:46 -04:00
daemons (~> 1.0, >= 1.0.9)
2015-11-25 11:18:44 -05:00
eventmachine (~> 1.0, >= 1.0.4)
rack (>= 1, < 3)
2017-02-21 12:01:53 -05:00
thor (0.19.4)
thread_safe (0.3.5)
tilt (2.0.5)
2016-05-21 21:17:15 -04:00
timecop (0.8.1)
timfel-krb5-auth (0.8.3)
2016-12-13 04:48:47 -05:00
tool (0.2.3)
truncato (0.7.8)
htmlentities (~> 4.3.1)
nokogiri (~> 1.6.1)
tzinfo (1.2.2)
2014-05-29 08:13:01 -04:00
thread_safe (~> 0.1)
u2f (0.2.1)
2015-10-14 02:39:59 -04:00
uglifier (2.7.2)
2011-10-08 17:36:38 -04:00
execjs (>= 0.3.0)
json (>= 1.8.0)
underscore-rails (1.8.3)
unf (0.1.4)
unf_ext
unf_ext (0.0.7.2)
2016-09-14 15:49:32 -04:00
unicode-display_width (1.1.1)
unicorn (5.1.0)
2013-07-08 02:47:31 -04:00
kgio (~> 2.6)
raindrops (~> 0.7)
2015-11-25 11:18:44 -05:00
unicorn-worker-killer (0.4.4)
2015-08-25 21:42:46 -04:00
get_process_mem (~> 0)
2015-11-25 11:18:44 -05:00
unicorn (>= 4, < 6)
uniform_notifier (1.10.0)
validates_hostname (1.0.6)
2016-02-09 12:06:55 -05:00
activerecord (>= 3.0)
activesupport (>= 3.0)
version_sorter (2.1.0)
2015-08-25 21:42:46 -04:00
virtus (1.0.5)
axiom-types (~> 0.1)
coercible (~> 1.0)
2015-08-25 21:42:46 -04:00
descendants_tracker (~> 0.0, >= 0.0.3)
equalizer (~> 0.0, >= 0.0.9)
vmstat (2.3.0)
warden (1.2.6)
2011-10-08 17:36:38 -04:00
rack (>= 1.0)
web-console (2.3.0)
2015-11-25 11:18:44 -05:00
activemodel (>= 4.0)
binding_of_caller (>= 0.7.2)
railties (>= 4.0)
sprockets-rails (>= 2.0, < 4.0)
webmock (1.21.0)
addressable (>= 2.3.6)
2013-05-01 06:29:29 -04:00
crack (>= 0.3.2)
webpack-rails (0.9.9)
rails (>= 3.2.0)
2015-11-25 11:18:44 -05:00
websocket-driver (0.6.3)
2015-06-18 22:14:41 -04:00
websocket-extensions (>= 0.1.0)
websocket-extensions (0.1.2)
2015-09-09 04:06:35 -04:00
wikicloth (0.8.1)
2014-07-28 05:47:27 -04:00
builder
expression_parser
2015-09-09 04:06:35 -04:00
rinku
xml-simple (1.1.5)
xpath (2.0.0)
2011-10-08 17:36:38 -04:00
nokogiri (~> 1.3)
PLATFORMS
ruby
DEPENDENCIES
RedCloth (~> 4.3.2)
ace-rails-ap (~> 4.1.0)
activerecord_sane_schema_dumper (= 0.2)
acts-as-taggable-on (~> 4.0)
2015-08-25 21:42:46 -04:00
addressable (~> 2.3.8)
after_commit_queue (~> 1.3.0)
akismet (~> 2.0)
allocations (~> 1.0)
2015-11-25 17:03:30 -05:00
asana (~> 0.4.0)
asciidoctor (~> 1.5.2)
asciidoctor-plantuml (= 0.0.7)
attr_encrypted (~> 3.0.0)
2015-08-25 21:42:46 -04:00
awesome_print (~> 1.2.0)
babosa (~> 1.0.2)
2016-05-02 07:29:17 -04:00
base32 (~> 0.3.0)
benchmark-ips (~> 2.3.0)
2015-08-25 21:42:46 -04:00
better_errors (~> 1.0.1)
binding_of_caller (~> 0.7.2)
2016-01-01 21:11:39 -05:00
bootstrap-sass (~> 3.3.0)
2017-01-18 12:21:57 -05:00
brakeman (~> 3.4.0)
browser (~> 2.2)
bullet (~> 5.2.0)
bundler-audit (~> 0.5.0)
2016-03-14 01:52:19 -04:00
capybara (~> 2.6.2)
2015-04-25 14:10:09 -04:00
capybara-screenshot (~> 1.0.0)
2016-03-03 20:32:18 -05:00
carrierwave (~> 0.10.0)
charlock_holmes (~> 0.7.3)
chronic (~> 0.10.2)
2016-05-18 16:21:51 -04:00
chronic_duration (~> 0.10.6)
2015-08-25 21:42:46 -04:00
coffee-rails (~> 4.1.0)
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
connection_pool (~> 2.0)
2015-11-24 15:42:42 -05:00
creole (~> 0.5.0)
2016-01-01 21:34:49 -05:00
d3_rails (~> 3.5.0)
database_cleaner (~> 1.5.0)
deckar01-task_list (= 1.0.6)
default_value_for (~> 3.0.0)
devise (~> 4.2)
devise-two-factor (~> 3.0.0)
diffy (~> 3.1.0)
doorkeeper (~> 4.2.0)
2015-08-25 21:42:46 -04:00
dropzonejs-rails (~> 0.7.1)
email_reply_trimmer (~> 0.1)
email_spec (~> 1.6.0)
2016-12-08 01:30:07 -05:00
factory_girl_rails (~> 4.7.0)
ffaker (~> 2.4)
flay (~> 2.6.1)
fog-aws (~> 0.9)
fog-core (~> 1.40)
fog-google (~> 0.5)
fog-local (~> 0.3)
fog-openstack (~> 0.1)
fog-rackspace (~> 0.1.1)
2016-06-13 08:05:08 -04:00
font-awesome-rails (~> 4.6.1)
foreman (~> 0.78.0)
2015-06-25 21:43:24 -04:00
fuubar (~> 2.0.0)
gemnasium-gitlab-service (~> 0.2)
gemojione (~> 3.0)
github-linguist (~> 4.7.0)
gitlab-flowdock-git-hook (~> 1.0.1)
gitlab-markup (~> 1.5.1)
2015-08-25 21:42:46 -04:00
gitlab_omniauth-ldap (~> 1.2.1)
gollum-lib (~> 4.2)
2016-02-28 07:11:43 -05:00
gollum-rugged_adapter (~> 0.4.2)
gon (~> 6.1.0)
google-api-client (~> 0.8.6)
2016-12-13 04:48:47 -05:00
grape (~> 0.18.0)
2016-11-21 03:09:19 -05:00
grape-entity (~> 0.6.0)
haml_lint (~> 0.18.2)
hamlit (~> 2.6.1)
health_check (~> 2.2.0)
2015-03-30 18:53:24 -04:00
hipchat (~> 1.5.0)
html-pipeline (~> 1.11.0)
html2text
2015-08-25 21:42:46 -04:00
httparty (~> 0.13.3)
Storing of application metrics in InfluxDB This adds the ability to write application metrics (e.g. SQL timings) to InfluxDB. These metrics can in turn be visualized using Grafana, or really anything else that can read from InfluxDB. These metrics can be used to track application performance over time, between different Ruby versions, different GitLab versions, etc. == Transaction Metrics Currently the following is tracked on a per transaction basis (a transaction is a Rails request or a single Sidekiq job): * Timings per query along with the raw (obfuscated) SQL and information about what file the query originated from. * Timings per view along with the path of the view and information about what file triggered the rendering process. * The duration of a request itself along with the controller/worker class and method name. * The duration of any instrumented method calls (more below). == Sampled Metrics Certain metrics can't be directly associated with a transaction. For example, a process' total memory usage is unrelated to any running transactions. While a transaction can result in the memory usage going up there's no accurate way to determine what transaction is to blame, this becomes especially problematic in multi-threaded environments. To solve this problem there's a separate thread that takes samples at a fixed interval. This thread (using the class Gitlab::Metrics::Sampler) currently tracks the following: * The process' total memory usage. * The number of file descriptors opened by the process. * The amount of Ruby objects (using ObjectSpace.count_objects). * GC statistics such as timings, heap slots, etc. The default/current interval is 15 seconds, any smaller interval might put too much pressure on InfluxDB (especially when running dozens of processes). == Method Instrumentation While currently not yet used methods can be instrumented to track how long they take to run. Unlike the likes of New Relic this doesn't require modifying the source code (e.g. including modules), it all happens from the outside. For example, to track `User.by_login` we'd add the following code somewhere in an initializer: Gitlab::Metrics::Instrumentation. instrument_method(User, :by_login) to instead instrument an instance method: Gitlab::Metrics::Instrumentation. instrument_instance_method(User, :save) Instrumentation for either all public model methods or a few crucial ones will be added in the near future, I simply haven't gotten to doing so just yet. == Configuration By default metrics are disabled. This means users don't have to bother setting anything up if they don't want to. Metrics can be enabled by editing one's gitlab.yml configuration file (see config/gitlab.yml.example for example settings). == Writing Data To InfluxDB Because InfluxDB is still a fairly young product I expect the worse. Data loss, unexpected reboots, the database not responding, you name it. Because of this data is _not_ written to InfluxDB directly, instead it's queued and processed by Sidekiq. This ensures that users won't notice anything when InfluxDB is giving trouble. The metrics worker can be started in a standalone manner as following: bundle exec sidekiq -q metrics The corresponding class is called MetricsWorker.
2015-12-09 10:45:51 -05:00
influxdb (~> 0.2)
2016-10-19 17:58:48 -04:00
jira-ruby (~> 1.1.2)
jquery-atwho-rails (~> 1.3.2)
jquery-rails (~> 4.1.0)
jquery-ui-rails (~> 5.0.0)
2016-08-01 13:38:46 -04:00
json-schema (~> 2.6.2)
jwt (~> 1.5.6)
kaminari (~> 0.17.0)
knapsack (~> 1.11.0)
kubeclient (~> 2.2.0)
letter_opener_web (~> 1.3.0)
license_finder (~> 2.1.0)
licensee (~> 8.7.0)
loofah (~> 2.0.3)
mail_room (~> 0.9.1)
method_source (~> 0.8)
2015-08-25 21:42:46 -04:00
minitest (~> 5.7.0)
mousetrap-rails (~> 1.4.6)
mysql2 (~> 0.3.16)
2015-11-25 11:18:44 -05:00
net-ssh (~> 3.0.1)
newrelic_rpm (~> 3.16)
2016-02-25 06:46:06 -05:00
nokogiri (~> 1.6.7, >= 1.6.7.2)
oauth2 (~> 1.2.0)
octokit (~> 4.6.2)
oj (~> 2.17.4)
omniauth (~> 1.3.2)
omniauth-auth0 (~> 1.4.1)
omniauth-authentiq (~> 0.3.0)
omniauth-azure-oauth2 (~> 0.0.6)
2015-11-11 23:25:31 -05:00
omniauth-cas3 (~> 1.1.2)
2016-09-13 07:09:04 -04:00
omniauth-facebook (~> 4.0.0)
2015-08-25 21:42:46 -04:00
omniauth-github (~> 1.1.1)
omniauth-gitlab (~> 1.0.2)
2016-06-23 10:15:10 -04:00
omniauth-google-oauth2 (~> 0.4.1)
2015-10-06 21:48:19 -04:00
omniauth-kerberos (~> 0.3.0)
omniauth-oauth2-generic (~> 0.2.2)
omniauth-saml (~> 1.7.0)
2015-10-06 22:03:42 -04:00
omniauth-shibboleth (~> 1.2.0)
2015-10-06 21:42:32 -04:00
omniauth-twitter (~> 1.2.0)
omniauth_crowd (~> 2.2.0)
2015-08-25 21:42:46 -04:00
org-ruby (~> 0.9.12)
paranoia (~> 2.2)
2015-08-25 21:42:46 -04:00
pg (~> 0.18.2)
2016-03-07 15:03:55 -05:00
poltergeist (~> 1.9.0)
premailer-rails (~> 1.9.0)
pry-byebug (~> 3.4.1)
pry-rails (~> 0.3.4)
rack-attack (~> 4.4.1)
2015-10-07 22:08:30 -04:00
rack-cors (~> 0.4.0)
2015-11-24 15:48:49 -05:00
rack-oauth2 (~> 1.2.1)
rack-proxy (~> 0.6.0)
2017-02-21 12:01:53 -05:00
rails (= 4.2.8)
2015-11-26 08:48:01 -05:00
rails-deprecated_sanitizer (~> 1.0.3)
rainbow (~> 2.1.0)
rblineprof (~> 0.3.6)
rdoc (~> 4.2)
recaptcha (~> 3.0)
redcarpet (~> 3.3.3)
redis (~> 3.2)
redis-namespace (~> 1.5.2)
redis-rails (~> 5.0.1)
request_store (~> 1.3)
2015-11-25 11:18:44 -05:00
responders (~> 2.0)
rouge (~> 2.0)
2015-08-25 21:42:46 -04:00
rqrcode-rails3 (~> 0.1.7)
rspec-rails (~> 3.5.0)
rspec-retry (~> 0.4.5)
rspec_profiling (~> 0.0.5)
rubocop (~> 0.46.0)
rubocop-rspec (~> 1.9.1)
2015-09-15 16:09:32 -04:00
ruby-fogbugz (~> 0.2.1)
ruby-prof (~> 0.16.2)
2017-01-04 13:43:06 -05:00
rugged (~> 0.24.0)
sanitize (~> 2.0)
sass-rails (~> 5.0.6)
scss_lint (~> 0.47.0)
2015-08-25 21:42:46 -04:00
seed-fu (~> 2.3.5)
2015-06-24 17:13:21 -04:00
select2-rails (~> 3.5.9)
sentry-raven (~> 2.0.0)
2015-08-25 21:42:46 -04:00
settingslogic (~> 2.0.9)
sham_rack (~> 1.3.6)
shoulda-matchers (~> 2.8.0)
sidekiq (~> 4.2.7)
sidekiq-cron (~> 0.4.4)
sidekiq-limit_fetch (~> 3.4)
simplecov (= 0.12.0)
2016-11-30 09:51:48 -05:00
slack-notifier (~> 1.5.1)
2015-08-25 21:42:46 -04:00
spinach-rails (~> 0.2.1)
2016-03-09 08:12:08 -05:00
spinach-rerun-reporter (~> 0.0.2)
spring (~> 1.7.0)
2015-08-25 21:42:46 -04:00
spring-commands-rspec (~> 1.0.4)
spring-commands-spinach (~> 1.1.0)
sprockets (~> 3.7.0)
stackprof (~> 0.2.10)
state_machines-activerecord (~> 0.4.0)
sys-filesystem (~> 1.1.6)
Fix race conditions for AuthorizedProjectsWorker There were two cases that could be problematic: 1. Because sometimes AuthorizedProjectsWorker would be scheduled in a transaction it was possible for a job to run/complete before a COMMIT; resulting in it either producing an error, or producing no new data. 2. When scheduling jobs the code would not wait until completion. This could lead to a user creating a project and then immediately trying to push to it. Usually this will work fine, but given enough load it might take a few seconds before a user has access. The first one is problematic, the second one is mostly just annoying (but annoying enough to warrant a solution). This commit changes two things to deal with this: 1. Sidekiq scheduling now takes places after a COMMIT, this is ensured by scheduling using Rails' after_commit hook instead of doing so in an arbitrary method. 2. When scheduling jobs the calling thread now waits for all jobs to complete. Solution 2 requires tracking of job completions. Sidekiq provides a way to find a job by its ID, but this involves scanning over the entire queue; something that is very in-efficient for large queues. As such a more efficient solution is necessary. There are two main Gems that can do this in a more efficient manner: * sidekiq-status * sidekiq_status No, this is not a joke. Both Gems do a similar thing (but slightly different), and the only difference in their name is a dash vs an underscore. Both Gems however provide far more than just checking if a job has been completed, and both have their problems. sidekiq-status does not appear to be actively maintained, with the last release being in 2015. It also has some issues during testing as API calls are not stubbed in any way. sidekiq_status on the other hand does not appear to be very popular, and introduces a similar amount of code. Because of this I opted to write a simple home grown solution. After all, all we need is storing a job ID somewhere so we can efficiently look it up; we don't need extra web UIs (as provided by sidekiq-status) or complex APIs to update progress, etc. This is where Gitlab::SidekiqStatus comes in handy. This namespace contains some code used for tracking, removing, and looking up job IDs; all without having to scan over an entire queue. Data is removed explicitly, but also expires automatically just in case. Using this API we can now schedule jobs in a fork-join like manner: we schedule the jobs in Sidekiq, process them in parallel, then wait for completion. By using Sidekiq we can leverage all the benefits such as being able to scale across multiple cores and hosts, retrying failed jobs, etc. The one downside is that we need to make sure we can deal with unexpected increases in job processing timings. To deal with this the class Gitlab::JobWaiter (used for waiting for jobs to complete) will only wait a number of seconds (30 by default). Once this timeout is reached it will simply return. For GitLab.com almost all AuthorizedProjectWorker jobs complete in seconds, only very rarely do we spike to job timings of around a minute. These in turn seem to be the result of external factors (e.g. deploys), in which case a user is most likely not able to use the system anyway. In short, this new solution should ensure that jobs are processed properly and that in almost all cases a user has access to their resources whenever they need to have access.
2017-01-22 12:22:02 -05:00
test_after_commit (~> 1.1)
thin (~> 1.7.0)
timecop (~> 0.8.0)
truncato (~> 0.7.8)
u2f (~> 0.2.1)
2015-10-14 02:39:59 -04:00
uglifier (~> 2.7.2)
underscore-rails (~> 1.8.0)
2015-08-25 21:42:46 -04:00
unf (~> 0.1.4)
unicorn (~> 5.1.0)
unicorn-worker-killer (~> 0.4.4)
validates_hostname (~> 1.0.6)
version_sorter (~> 2.1.0)
2015-08-25 21:42:46 -04:00
virtus (~> 1.0.1)
vmstat (~> 2.3.0)
2015-11-25 11:18:44 -05:00
web-console (~> 2.0)
webmock (~> 1.21.0)
webpack-rails (~> 0.9.9)
2015-09-09 04:06:35 -04:00
wikicloth (= 0.8.1)
BUNDLED WITH
1.14.3