X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/87f17f28080eef37ad940a039fd1623e72e3058e..0474aa89b8f497fd5abaa072f9bc501742cc5b42:/services/api/lib/eventbus.rb

diff --git a/services/api/lib/eventbus.rb b/services/api/lib/eventbus.rb
index c93f2e5f4b..03afc31601 100644
--- a/services/api/lib/eventbus.rb
+++ b/services/api/lib/eventbus.rb
@@ -4,6 +4,7 @@ require 'faye/websocket'
 require 'record_filters'
 require 'load_param'
 
+# Patch in user, last_log_id and filters fields into the Faye::Websocket class.
 module Faye
   class WebSocket
     attr_accessor :user
@@ -12,108 +13,191 @@ module Faye
   end
 end
 
+# Store the filters supplied by the user that will be applied to the logs table
+# to determine which events to return to the listener.
 class Filter
   include LoadParam
 
+  attr_accessor :filters
+
   def initialize p
-    @p = p
+    @params = p
     load_filters_param
   end
 
   def params
-    @p
-  end
-
-  def filters
-    @filters
+    @params
   end
 end
 
+# Manages websocket connections, accepts subscription messages and publishes
+# log table events.
 class EventBus
   include CurrentApiClient
   include RecordFilters
 
+  # used in RecordFilters
+  def model_class
+    Log
+  end
+
+  # Initialize EventBus.  Takes no parameters.
   def initialize
     @channel = EventMachine::Channel.new
     @mtx = Mutex.new
     @bgthread = false
   end
 
-  def on_connect ws
-    if not current_user
-      ws.send ({status: 401, message: "Valid API token required"}.to_json)
-      ws.close
-      return
-    end
-
-    ws.user = current_user
-    ws.filters = []
-    ws.last_log_id = nil
-
-    sub = @channel.subscribe do |msg|
+  # Push out any pending events to the connection +ws+
+  # +id+  the id of the most recent row in the log table, may be nil
+  def push_events ws, id = nil
       begin
-      puts "Waking up"
+        # Must have at least one filter set up to receive events
+        if ws.filters.length > 0
+          # Start with log rows readable by user, sorted in ascending order
+          logs = Log.readable_by(ws.user).order("id asc")
+
+          cond_id = nil
+          cond_out = []
+          param_out = []
+
+          if ws.last_log_id
+            # Client is only interested in log rows that are newer than the
+            # last log row seen by the client.
+            cond_id = "logs.id > ?"
+            param_out << ws.last_log_id
+          elsif id
+            # No last log id, so only look at the most recently changed row
+            cond_id = "logs.id = ?"
+            param_out << id.to_i
+          else
+            return
+          end
 
-      # Must have at least one filter set up to receive events
-      if ws.filters.length > 0
+          # Now process filters provided by client
+          ws.filters.each do |filter|
+            ft = record_filters filter.filters, Log
+            if ft[:cond_out].any?
+              # Join the clauses within a single subscription filter with AND
+              # so it is consistent with regular queries
+              cond_out << "(#{ft[:cond_out].join ') AND ('})"
+              param_out += ft[:param_out]
+            end
+          end
 
-        # Start with log rows readable by user, sorted in ascending order
-        logs = Log.readable_by(ws.user).order("id asc")
+          # Add filters to query
+          if cond_out.any?
+            # Join subscriptions with OR
+            logs = logs.where(cond_id + " AND ((#{cond_out.join ') OR ('}))", *param_out)
+          else
+            logs = logs.where(cond_id, *param_out)
+          end
 
-        if ws.last_log_id
-          # Only get log rows that are new
-          logs = logs.where("logs.id > ? and logs.id <= ?", ws.last_log_id, msg.to_i)
-        else
-          # No last log id, so only look at the most recently changed row
-          logs = logs.where("logs.id = ?", msg.to_i)
+          # Finally execute query and actually send the matching log rows
+          logs.each do |l|
+            ws.send(l.as_api_response.to_json)
+            ws.last_log_id = l.id
+          end
+        elsif id
+          # No filters set up, so just record the sequence number
+          ws.last_log_id = id.to_i
         end
+      rescue Exception => e
+        Rails.logger.warn "Error publishing event: #{$!}"
+        Rails.logger.warn "Backtrace:\n\t#{e.backtrace.join("\n\t")}"
+        ws.send ({status: 500, message: 'error'}.to_json)
+        ws.close
+      end
+  end
+
+  # Handle inbound subscribe or unsubscribe message.
+  def handle_message ws, event
+    begin
+      # Parse event data as JSON
+      p = (Oj.load event.data).symbolize_keys
 
-        # Record the most recent row
-        ws.last_log_id = msg.to_i
+      if p[:method] == 'subscribe'
+        # Handle subscribe event
 
-        # Now process filters provided by client
-        cond_out = []
-        param_out = []
-        ws.filters.each do |filter|
-          ft = record_filters filter.filters
-          cond_out += ft[:cond_out]
-          param_out += ft[:param_out]
+        if p[:last_log_id]
+          # Set or reset the last_log_id.  The event bus only reports events
+          # for rows that come after last_log_id.
+          ws.last_log_id = p[:last_log_id].to_i
         end
 
-        # Add filters to query
-        if cond_out.any?
-          logs = logs.where(cond_out.join(' OR '), *param_out)
+        if ws.filters.length < MAX_FILTERS
+          # Add a filter.  This gets the :filters field which is the same
+          # format as used for regular index queries.
+          ws.filters << Filter.new(p)
+          ws.send ({status: 200, message: 'subscribe ok'}.to_json)
+
+          # Send any pending events
+          push_events ws
+        else
+          ws.send ({status: 403, message: "maximum of #{MAX_FILTERS} filters allowed per connection"}.to_json)
         end
 
-        # Finally execute query and send matching rows
-        logs.each do |l|
-          ws.send(l.as_api_response.to_json)
+      elsif p[:method] == 'unsubscribe'
+        # Handle unsubscribe event
+
+        len = ws.filters.length
+        ws.filters.select! { |f| not ((f.filters == p[:filters]) or (f.filters.empty? and p[:filters].nil?)) }
+        if ws.filters.length < len
+          ws.send ({status: 200, message: 'unsubscribe ok'}.to_json)
+        else
+          ws.send ({status: 404, message: 'filter not found'}.to_json)
         end
+
       else
-        # No filters set up, so just record the sequence number
-        ws.last_log_id.nil = msg.to_i
-      end
-      rescue Exception => e
-        puts "#{e}"
+        ws.send ({status: 400, message: "missing or unrecognized method"}.to_json)
       end
+    rescue Oj::Error => e
+      ws.send ({status: 400, message: "malformed request"}.to_json)
+    rescue Exception => e
+      Rails.logger.warn "Error handling message: #{$!}"
+      Rails.logger.warn "Backtrace:\n\t#{e.backtrace.join("\n\t")}"
+      ws.send ({status: 500, message: 'error'}.to_json)
+      ws.close
     end
+  end
+
+  # Constant maximum number of filters, to avoid silly huge database queries.
+  MAX_FILTERS = 16
+
+  # Called by RackSocket when a new websocket connection has been established.
+  def on_connect ws
 
+    # Disconnect if no valid API token.
+    # current_user is included from CurrentApiClient
+    if not current_user
+      ws.send ({status: 401, message: "Valid API token required"}.to_json)
+      ws.close
+      return
+    end
+
+    # Initialize our custom fields on the websocket connection object.
+    ws.user = current_user
+    ws.filters = []
+    ws.last_log_id = nil
+
+    # Subscribe to internal postgres notifications through @channel.  This will
+    # call push_events when a notification comes through.
+    sub = @channel.subscribe do |msg|
+      push_events ws, msg
+    end
+
+    # Set up callback for inbound message dispatch.
     ws.on :message do |event|
-      p = Oj.load event.data
-      if p["method"] == 'subscribe'
-        if p["starting_log_id"]
-          ws.last_log_id = p["starting_log_id"].to_i
-        end
-        ws.filters.push(Filter.new p)
-        ws.send ({status: 200, message: 'subscribe ok'}.to_json)
-      end
+      handle_message ws, event
     end
 
+    # Set up socket close callback
     ws.on :close do |event|
       @channel.unsubscribe sub
       ws = nil
     end
 
+    # Start up thread to monitor the Postgres database, if none exists already.
     @mtx.synchronize do
       unless @bgthread
         @bgthread = true
@@ -124,6 +208,12 @@ class EventBus
             begin
               conn.async_exec "LISTEN logs"
               while true
+                # wait_for_notify will block until there is a change
+                # notification from Postgres about the logs table, then push
+                # the notification into the EventMachine channel.  Each
+                # websocket connection subscribes to the other end of the
+                # channel and calls #push_events to actually dispatch the
+                # events to the client.
                 conn.wait_for_notify do |channel, pid, payload|
                   @channel.push payload
                 end
@@ -135,8 +225,13 @@ class EventBus
               conn.async_exec "UNLISTEN *"
             end
           end
+          @bgthread = false
         end
       end
     end
+
+    # Since EventMachine is an asynchronous event based dispatcher, #on_connect
+    # does not block but instead returns immediately after having set up the
+    # websocket and notification channel callbacks.
   end
 end