Move the circuitbreaker check out in a separate process

Moving the check out of the general requests, makes sure we don't have
any slowdown in the regular requests.

To keep the process performing this checks small, the check is still
performed inside a unicorn. But that is called from a process running
on the same server.

Because the checks are now done outside normal request, we can have a
simpler failure strategy:

The check is now performed in the background every
`circuitbreaker_check_interval`. Failures are logged in redis. The
failures are reset when the check succeeds. Per check we will try
`circuitbreaker_access_retries` times within
`circuitbreaker_storage_timeout` seconds.

When the number of failures exceeds
`circuitbreaker_failure_count_threshold`, we will block access to the
storage.

After `failure_reset_time` of no checks, we will clear the stored
failures. This could happen when the process that performs the checks
is not running.

4 files changed, 83 insertions, 4 deletions

diff --git a/db/migrate/20171123094802_add_circuitbreaker_check_interval_to_application_settings.rb b/db/migrate/20171123094802_add_circuitbreaker_check_interval_to_application_settings.rb
new file mode 100644
index 00000000000..213d46018fc
--- /dev/null
+++ b/db/migrate/20171123094802_add_circuitbreaker_check_interval_to_application_settings.rb
@@ -0,0 +1,20 @@
+class AddCircuitbreakerCheckIntervalToApplicationSettings < ActiveRecord::Migration
+  include Gitlab::Database::MigrationHelpers
+
+  # Set this constant to true if this migration requires downtime.
+  DOWNTIME = false
+
+  disable_ddl_transaction!
+
+  def up
+    add_column_with_default :application_settings,
+                            :circuitbreaker_check_interval,
+                            :integer,
+                            default: 1
+  end
+
+  def down
+    remove_column :application_settings,
+                  :circuitbreaker_check_interval
+  end
+end
diff --git a/db/post_migrate/20171123101020_update_circuitbreaker_defaults.rb b/db/post_migrate/20171123101020_update_circuitbreaker_defaults.rb
new file mode 100644
index 00000000000..8e1c9e6d6bb
--- /dev/null
+++ b/db/post_migrate/20171123101020_update_circuitbreaker_defaults.rb
@@ -0,0 +1,34 @@
+# See http://doc.gitlab.com/ce/development/migration_style_guide.html
+# for more information on how to write migrations for GitLab.
+
+class UpdateCircuitbreakerDefaults < ActiveRecord::Migration
+  include Gitlab::Database::MigrationHelpers
+
+  DOWNTIME = false
+
+  class ApplicationSetting < ActiveRecord::Base; end
+
+  def up
+    change_column_default :application_settings,
+                          :circuitbreaker_failure_count_threshold,
+                          3
+    change_column_default :application_settings,
+                          :circuitbreaker_storage_timeout,
+                          15
+
+    ApplicationSetting.update_all(circuitbreaker_failure_count_threshold: 3,
+                                  circuitbreaker_storage_timeout: 15)
+  end
+
+  def down
+    change_column_default :application_settings,
+                          :circuitbreaker_failure_count_threshold,
+                          160
+    change_column_default :application_settings,
+                          :circuitbreaker_storage_timeout,
+                          30
+
+    ApplicationSetting.update_all(circuitbreaker_failure_count_threshold: 160,
+                                  circuitbreaker_storage_timeout: 30)
+  end
+end
diff --git a/db/post_migrate/20171123101046_remove_old_circuitbreaker_config.rb b/db/post_migrate/20171123101046_remove_old_circuitbreaker_config.rb
new file mode 100644
index 00000000000..e646d4d3224
--- /dev/null
+++ b/db/post_migrate/20171123101046_remove_old_circuitbreaker_config.rb
@@ -0,0 +1,26 @@
+# See http://doc.gitlab.com/ce/development/migration_style_guide.html
+# for more information on how to write migrations for GitLab.
+
+class RemoveOldCircuitbreakerConfig < ActiveRecord::Migration
+  include Gitlab::Database::MigrationHelpers
+
+  DOWNTIME = false
+
+  def up
+    remove_column :application_settings,
+                  :circuitbreaker_backoff_threshold
+    remove_column :application_settings,
+                  :circuitbreaker_failure_wait_time
+  end
+
+  def down
+    add_column :application_settings,
+               :circuitbreaker_backoff_threshold,
+               :integer,
+               default: 80
+    add_column :application_settings,
+               :circuitbreaker_failure_wait_time,
+               :integer,
+               default: 30
+  end
+end
diff --git a/db/schema.rb b/db/schema.rb
index 6ea3ab54742..4c697a4a384 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -135,12 +135,10 @@ ActiveRecord::Schema.define(version: 20171205190711) do
     t.boolean "hashed_storage_enabled", default: false, null: false
     t.boolean "project_export_enabled", default: true, null: false
     t.boolean "auto_devops_enabled", default: false, null: false
-    t.integer "circuitbreaker_failure_count_threshold", default: 160
-    t.integer "circuitbreaker_failure_wait_time", default: 30
+    t.integer "circuitbreaker_failure_count_threshold", default: 3
     t.integer "circuitbreaker_failure_reset_time", default: 1800
-    t.integer "circuitbreaker_storage_timeout", default: 30
+    t.integer "circuitbreaker_storage_timeout", default: 15
     t.integer "circuitbreaker_access_retries", default: 3
-    t.integer "circuitbreaker_backoff_threshold", default: 80
     t.boolean "throttle_unauthenticated_enabled", default: false, null: false
     t.integer "throttle_unauthenticated_requests_per_period", default: 3600, null: false
     t.integer "throttle_unauthenticated_period_in_seconds", default: 3600, null: false
@@ -150,6 +148,7 @@ ActiveRecord::Schema.define(version: 20171205190711) do
     t.boolean "throttle_authenticated_web_enabled", default: false, null: false
     t.integer "throttle_authenticated_web_requests_per_period", default: 7200, null: false
     t.integer "throttle_authenticated_web_period_in_seconds", default: 3600, null: false
+    t.integer "circuitbreaker_check_interval", default: 1, null: false
     t.boolean "password_authentication_enabled_for_web"
     t.boolean "password_authentication_enabled_for_git", default: true
     t.integer "gitaly_timeout_default", default: 55, null: false