multi: unify channel info request and expose flap count

When we list channels, we get lifetime, uptime and now flap count from
the channel event store. This call unifies the values that external
callers will want (rather than splitting out a third request type just
for flaps) and exposes flap count over rpc.

We hardcode calculating uptime for the channel's whole life because
the listChannels endpoint doesn't start start/end params. Since we plan
on adding a separate uptime endpoint once we have persisted uptime, it's
ok to remove these parameters.

chanfitness/chaneventstore.go

@@ -47,11 +47,8 @@ type ChannelEventStore struct {
 	// and offline events.
 	peers map[route.Vertex]bool
 
-	// lifespanRequests serves requests for the lifespan of channels.
-	lifespanRequests chan lifespanRequest
-
-	// uptimeRequests serves requests for the uptime of channels.
-	uptimeRequests chan uptimeRequest
+	// chanInfoRequests serves requests for information about our channel.
+	chanInfoRequests chan channelInfoRequest
 
 	quit chan struct{}
 
@@ -94,36 +91,16 @@ type Config struct {
 	FlapCountTicker ticker.Ticker
 }
 
-// lifespanRequest contains the channel ID required to query the store for a
-// channel's lifespan and a blocking response channel on which the result is
-// sent.
-type lifespanRequest struct {
-	channelPoint wire.OutPoint
-	responseChan chan lifespanResponse
-}
-
-// lifespanResponse contains the response to a lifespanRequest and an error if
-// one occurred.
-type lifespanResponse struct {
-	start time.Time
-	end   time.Time
-	err   error
-}
-
-// uptimeRequest contains the parameters required to query the store for a
-// channel's uptime and a blocking response channel on which the result is sent.
-type uptimeRequest struct {
+type channelInfoRequest struct {
 	channelPoint wire.OutPoint
-	startTime    time.Time
-	endTime      time.Time
-	responseChan chan uptimeResponse
+	start        time.Time
+	end          time.Time
+	responseChan chan channelInfoResponse
 }
 
-// uptimeResponse contains the response to an uptimeRequest and an error if one
-// occurred.
-type uptimeResponse struct {
-	uptime time.Duration
-	err    error
+type channelInfoResponse struct {
+	info *ChannelInfo
+	err  error
 }
 
 // NewChannelEventStore initializes an event store with the config provided.
@@ -134,8 +111,7 @@ func NewChannelEventStore(config *Config) *ChannelEventStore {
 		cfg:              config,
 		channels:         make(map[wire.OutPoint]*chanEventLog),
 		peers:            make(map[route.Vertex]bool),
-		lifespanRequests: make(chan lifespanRequest),
-		uptimeRequests:   make(chan uptimeRequest),
+		chanInfoRequests: make(chan channelInfoRequest),
 		quit:             make(chan struct{}),
 	}
 
@@ -365,35 +341,10 @@ func (c *ChannelEventStore) consume(subscriptions *subscriptions) {
 			}
 
 		// Serve all requests for channel lifetime.
-		case req := <-c.lifespanRequests:
-			var resp lifespanResponse
-
-			channel, ok := c.channels[req.channelPoint]
-			if !ok {
-				resp.err = ErrChannelNotFound
-			} else {
-				resp.start = channel.openedAt
-				resp.end = channel.closedAt
-			}
-
-			req.responseChan <- resp
-
-		// Serve requests for channel uptime.
-		case req := <-c.uptimeRequests:
-			var resp uptimeResponse
-
-			channel, ok := c.channels[req.channelPoint]
-			if !ok {
-				resp.err = ErrChannelNotFound
-			} else {
-				uptime, err := channel.uptime(
-					req.startTime, req.endTime,
-				)
-
-				resp.uptime = uptime
-				resp.err = err
-			}
+		case req := <-c.chanInfoRequests:
+			var resp channelInfoResponse
 
+			resp.info, resp.err = c.getChanInfo(req)
 			req.responseChan <- resp
 
 		case <-c.cfg.FlapCountTicker.Ticks():
@@ -409,67 +360,83 @@ func (c *ChannelEventStore) consume(subscriptions *subscriptions) {
 	}
 }
 
-// GetLifespan returns the opening and closing time observed for a channel and
-// a boolean to indicate whether the channel is known the the event store. If
-// the channel is still open, a zero close time is returned.
-func (c *ChannelEventStore) GetLifespan(
-	channelPoint wire.OutPoint) (time.Time, time.Time, error) {
+// ChannelInfo provides the set of information that the event store has recorded
+// for a channel.
+type ChannelInfo struct {
+	// Lifetime is the total amount of time we have monitored the channel
+	// for.
+	Lifetime time.Duration
+
+	// Uptime is the total amount of time that the channel peer has been
+	// observed as online during the monitored lifespan.
+	Uptime time.Duration
+
+	// FlapCount is the number of times the channel peer has gone offline
+	// over our record of the peer. Note that these flaps are tracked across
+	// channels (some of these flaps may have been before this channel was
+	// opened).
+	FlapCount uint32
+}
 
-	request := lifespanRequest{
+// GetChanInfo gets all the information we have on a channel in the event store.
+func (c *ChannelEventStore) GetChanInfo(channelPoint wire.OutPoint) (*ChannelInfo,
+	error) {
+
+	request := channelInfoRequest{
 		channelPoint: channelPoint,
-		responseChan: make(chan lifespanResponse),
+		start:        time.Time{},
+		end:          time.Time{},
+		responseChan: make(chan channelInfoResponse),
 	}
 
-	// Send a request for the channel's lifespan to the main event loop, or
-	// return early with an error if the store has already received a
+	// Send a request for the channel's information to the main event loop,
+	// or return early with an error if the store has already received a
 	// shutdown signal.
 	select {
-	case c.lifespanRequests <- request:
+	case c.chanInfoRequests <- request:
 	case <-c.quit:
-		return time.Time{}, time.Time{}, errShuttingDown
+		return nil, errShuttingDown
 	}
 
 	// Return the response we receive on the response channel or exit early
 	// if the store is instructed to exit.
 	select {
 	case resp := <-request.responseChan:
-		return resp.start, resp.end, resp.err
+		return resp.info, resp.err
 
 	case <-c.quit:
-		return time.Time{}, time.Time{}, errShuttingDown
+		return nil, errShuttingDown
 	}
 }
 
-// GetUptime returns the uptime of a channel over a period and an error if the
-// channel cannot be found or the uptime calculation fails.
-func (c *ChannelEventStore) GetUptime(channelPoint wire.OutPoint, startTime,
-	endTime time.Time) (time.Duration, error) {
+// getChanInfo collects channel information for a channel. It gets uptime over
+// the full life of the channel.
+func (c *ChannelEventStore) getChanInfo(req channelInfoRequest) (*ChannelInfo,
+	error) {
 
-	request := uptimeRequest{
-		channelPoint: channelPoint,
-		startTime:    startTime,
-		endTime:      endTime,
-		responseChan: make(chan uptimeResponse),
+	// Look for the channel in our current set.
+	channel, ok := c.channels[req.channelPoint]
+	if !ok {
+		return nil, ErrChannelNotFound
 	}
 
-	// Send a request for the channel's uptime to the main event loop, or
-	// return early with an error if the store has already received a
-	// shutdown signal.
-	select {
-	case c.uptimeRequests <- request:
-	case <-c.quit:
-		return 0, errShuttingDown
+	// If our channel is not closed, we want to calculate uptime until the
+	// present.
+	endTime := channel.closedAt
+	if endTime.IsZero() {
+		endTime = c.cfg.Clock.Now()
 	}
 
-	// Return the response we receive on the response channel or exit early
-	// if the store is instructed to exit.
-	select {
-	case resp := <-request.responseChan:
-		return resp.uptime, resp.err
-
-	case <-c.quit:
-		return 0, errShuttingDown
+	uptime, err := channel.uptime(channel.openedAt, endTime)
+	if err != nil {
+		return nil, err
 	}
+
+	return &ChannelInfo{
+		Lifetime:  endTime.Sub(channel.openedAt),
+		Uptime:    uptime,
+		FlapCount: channel.flapCount,
+	}, nil
 }
 
 // recordFlapCount will record our flap count for each peer that we currently