Class: Mongo::Cluster::CursorReaper Private

Inherits:
Object
  • Object
show all
Includes:
Retryable
Defined in:
lib/mongo/cluster/reapers/cursor_reaper.rb

Overview

This class is part of a private API. You should avoid using this class if possible, as it may be removed or be changed in the future.

A manager that sends kill cursors operations at regular intervals to close cursors that have been garbage collected without being exhausted.

Since:

  • 2.3.0

Constant Summary collapse

FREQUENCY =

This constant is part of a private API. You should avoid using this constant if possible, as it may be removed or be changed in the future.

The default time interval for the cursor reaper to send pending kill cursors operations.

Since:

  • 2.3.0

1

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from Retryable

#read_worker, #select_server, #with_overload_retry, #write_worker

Constructor Details

#initialize(cluster) ⇒ CursorReaper

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Create a cursor reaper.

Parameters:

  • cluster (Cluster) —

    The cluster.

Since:

  • 2.3.0



39
40
41
42
43
44
45
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 39

def initialize(cluster)
  @cluster = cluster
  @to_kill = {}
  @active_cursor_ids = Set.new
  @mutex = Mutex.new
  @kill_spec_queue = Queue.new
end

Instance Attribute Details

#cluster ⇒ Object (readonly)

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Since:

  • 2.3.0



47
48
49
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 47

def cluster
  @cluster
end

Instance Method Details

#kill_cursors ⇒ Object Also known as: execute, flush

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Execute all pending kill cursors operations.

Examples:

Execute pending kill cursors operations.

cursor_reaper.kill_cursors

Since:

  • 2.3.0



126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 126

def kill_cursors
  # TODO: optimize this to batch kill cursor operations for the same
  # server/database/collection instead of killing each cursor
  # individually.
  loop do
    server_address = nil

    kill_spec = @mutex.synchronize do
      read_scheduled_kill_specs
      # Find a server that has any cursors scheduled for destruction.
      server_address, specs =
        @to_kill.detect { |_, specs| specs.any? }

      if specs.nil?
        # All servers have empty specs, nothing to do.
        return
      end

      # Note that this mutates the spec in the queue.
      # If the kill cursor operation fails, we don't attempt to
      # kill that cursor again.
      spec = specs.take(1).tap do |arr|
        specs.subtract(arr)
      end.first

      unless @active_cursor_ids.include?(spec.cursor_id)
        # The cursor was already killed, typically because it has
        # been iterated to completion. Remove the kill spec from
        # our records without doing any more work.
        spec = nil
      end

      spec
    end

    # If there was a spec to kill but its cursor was already killed,
    # look for another spec.
    next unless kill_spec

    # We could also pass kill_spec directly into the KillCursors
    # operation, though this would make that operation have a
    # different API from all of the other ones which accept hashes.
    spec = {
      cursor_ids: [ kill_spec.cursor_id ],
      coll_name: kill_spec.coll_name,
      db_name: kill_spec.db_name,
    }
    op = Operation::KillCursors.new(spec)

    server = cluster.servers.detect do |server|
      server.address == server_address
    end

    unless server
      # The server for this cursor has gone away --- maybe temporarily,
      # maybe permanently, but we can't know. To prevent connections from
      # leaking in the case of a permanent failure, we'll just silently
      # drop this killspec and move on.
      next
    end

    options = {
      server_api: server.options[:server_api],
      connection_global_id: kill_spec.connection_global_id,
    }
    if connection = kill_spec.connection
      op.execute_with_connection(connection, context: Operation::Context.new(options: options))
      connection.connection_pool.check_in(connection)
    else
      op.execute(server, context: Operation::Context.new(options: options))
    end

    next unless session = kill_spec.session

    session.end_session if session.implicit?
  end
end

#read_scheduled_kill_specs ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Read and decode scheduled kill cursors operations.

This method mutates instance variables without locking, so it is not thread safe. Generally, it should not be called itself, this is a helper for ‘kill_cursor` method.

Since:

  • 2.3.0



103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 103

def read_scheduled_kill_specs
  while kill_spec = @kill_spec_queue.pop(true)
    if @active_cursor_ids.include?(kill_spec.cursor_id)
      @to_kill[kill_spec.server_address] ||= Set.new
      @to_kill[kill_spec.server_address] << kill_spec
    elsif (session = kill_spec.session) && session.implicit?
      # Cursor was already closed; end the session immediately to release
      # references rather than waiting for the kill_spec to go out of scope.
      session.end_session
    end
  end
rescue ThreadError
  # Empty queue, nothing to do.
end

#register_cursor(id) ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Register a cursor id as active.

Examples:

Register a cursor as active.

cursor_reaper.register_cursor(id)

Parameters:

  • id (Integer) —

    The id of the cursor to register as active.

Raises:

  • (ArgumentError)

Since:

  • 2.3.0



68
69
70
71
72
73
74
75
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 68

def register_cursor(id)
  raise ArgumentError, 'register_cursor called with nil cursor_id' if id.nil?
  raise ArgumentError, 'register_cursor called with cursor_id=0' if id == 0

  @mutex.synchronize do
    @active_cursor_ids << id
  end
end

#schedule_kill_cursor(kill_spec) ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Schedule a kill cursors operation to be eventually executed.

Parameters:

Since:

  • 2.3.0



54
55
56
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 54

def schedule_kill_cursor(kill_spec)
  @kill_spec_queue << kill_spec
end

#unregister_cursor(id) ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Unregister a cursor id, indicating that it’s no longer active.

Examples:

Unregister a cursor.

cursor_reaper.unregister_cursor(id)

Parameters:

  • id (Integer) —

    The id of the cursor to unregister.

Raises:

  • (ArgumentError)

Since:

  • 2.3.0



87
88
89
90
91
92
93
94
 
# File 'lib/mongo/cluster/reapers/cursor_reaper.rb', line 87

def unregister_cursor(id)
  raise ArgumentError, 'unregister_cursor called with nil cursor_id' if id.nil?
  raise ArgumentError, 'unregister_cursor called with cursor_id=0' if id == 0

  @mutex.synchronize do
    @active_cursor_ids.delete(id)
  end
end