Merge pull request #1732 from fach/master

Adding backlog/current queue length to qdisc collector

8 files changed, 105 insertions, 13 deletions

diff --git a/collector/fixtures/e2e-64k-page-output.txt b/collector/fixtures/e2e-64k-page-output.txt
index fe32915..7b857ff 100644
--- a/collector/fixtures/e2e-64k-page-output.txt
+++ b/collector/fixtures/e2e-64k-page-output.txt
@@ -2497,10 +2497,18 @@ node_procs_blocked 0
 # HELP node_procs_running Number of processes in runnable state.
 # TYPE node_procs_running gauge
 node_procs_running 2
+# HELP node_qdisc_backlog Number of bytes currently in queue to be sent.
+# TYPE node_qdisc_backlog gauge
+node_qdisc_backlog{device="eth0",kind="pfifo_fast"} 0
+node_qdisc_backlog{device="wlan0",kind="fq"} 0
 # HELP node_qdisc_bytes_total Number of bytes sent.
 # TYPE node_qdisc_bytes_total counter
 node_qdisc_bytes_total{device="eth0",kind="pfifo_fast"} 83
 node_qdisc_bytes_total{device="wlan0",kind="fq"} 42
+# HELP node_qdisc_current_queue_length Number of packets currently in queue to be sent.
+# TYPE node_qdisc_current_queue_length gauge
+node_qdisc_current_queue_length{device="eth0",kind="pfifo_fast"} 0
+node_qdisc_current_queue_length{device="wlan0",kind="fq"} 0
 # HELP node_qdisc_drops_total Number of packets dropped.
 # TYPE node_qdisc_drops_total counter
 node_qdisc_drops_total{device="eth0",kind="pfifo_fast"} 0
diff --git a/collector/fixtures/e2e-output.txt b/collector/fixtures/e2e-output.txt
index 0edc65b..e8a5779 100644
--- a/collector/fixtures/e2e-output.txt
+++ b/collector/fixtures/e2e-output.txt
@@ -2566,10 +2566,18 @@ node_procs_blocked 0
 # HELP node_procs_running Number of processes in runnable state.
 # TYPE node_procs_running gauge
 node_procs_running 2
+# HELP node_qdisc_backlog Number of bytes currently in queue to be sent.
+# TYPE node_qdisc_backlog gauge
+node_qdisc_backlog{device="eth0",kind="pfifo_fast"} 0
+node_qdisc_backlog{device="wlan0",kind="fq"} 0
 # HELP node_qdisc_bytes_total Number of bytes sent.
 # TYPE node_qdisc_bytes_total counter
 node_qdisc_bytes_total{device="eth0",kind="pfifo_fast"} 83
 node_qdisc_bytes_total{device="wlan0",kind="fq"} 42
+# HELP node_qdisc_current_queue_length Number of packets currently in queue to be sent.
+# TYPE node_qdisc_current_queue_length gauge
+node_qdisc_current_queue_length{device="eth0",kind="pfifo_fast"} 0
+node_qdisc_current_queue_length{device="wlan0",kind="fq"} 0
 # HELP node_qdisc_drops_total Number of packets dropped.
 # TYPE node_qdisc_drops_total counter
 node_qdisc_drops_total{device="eth0",kind="pfifo_fast"} 0
diff --git a/collector/qdisc_linux.go b/collector/qdisc_linux.go
index daba199..ba462a0 100644
--- a/collector/qdisc_linux.go
+++ b/collector/qdisc_linux.go
@@ -32,6 +32,8 @@ type qdiscStatCollector struct {
         drops      typedDesc
         requeues   typedDesc
         overlimits typedDesc
+        qlength    typedDesc
+        backlog    typedDesc
         logger     log.Logger
 }
 
@@ -71,6 +73,16 @@ func NewQdiscStatCollector(logger log.Logger) (Collector, error) {
                         "Number of overlimit packets.",
                         []string{"device", "kind"}, nil,
                 ), prometheus.CounterValue},
+                qlength: typedDesc{prometheus.NewDesc(
+                        prometheus.BuildFQName(namespace, "qdisc", "current_queue_length"),
+                        "Number of packets currently in queue to be sent.",
+                        []string{"device", "kind"}, nil,
+                ), prometheus.GaugeValue},
+                backlog: typedDesc{prometheus.NewDesc(
+                        prometheus.BuildFQName(namespace, "qdisc", "backlog"),
+                        "Number of bytes currently in queue to be sent.",
+                        []string{"device", "kind"}, nil,
+                ), prometheus.GaugeValue},
                 logger: logger,
         }, nil
 }
@@ -114,6 +126,8 @@ func (c *qdiscStatCollector) Update(ch chan<- prometheus.Metric) error {
                 ch <- c.drops.mustNewConstMetric(float64(msg.Drops), msg.IfaceName, msg.Kind)
                 ch <- c.requeues.mustNewConstMetric(float64(msg.Requeues), msg.IfaceName, msg.Kind)
                 ch <- c.overlimits.mustNewConstMetric(float64(msg.Overlimits), msg.IfaceName, msg.Kind)
+                ch <- c.qlength.mustNewConstMetric(float64(msg.Qlen), msg.IfaceName, msg.Kind)
+                ch <- c.backlog.mustNewConstMetric(float64(msg.Backlog), msg.IfaceName, msg.Kind)
         }
 
         return nil
diff --git a/go.mod b/go.mod
index 9a92e3b..a98b7db 100644
--- a/go.mod
+++ b/go.mod
@@ -4,7 +4,7 @@ require (
         github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d // indirect
         github.com/beevik/ntp v0.3.0
         github.com/coreos/go-systemd v0.0.0-20191104093116-d3cd4ed1dbcf
-        github.com/ema/qdisc v0.0.0-20190904071900-b82c76788043
+        github.com/ema/qdisc v0.0.0-20200603082823-62d0308e3e00
         github.com/go-kit/kit v0.10.0
         github.com/godbus/dbus v0.0.0-20190402143921-271e53dc4968
         github.com/golang/protobuf v1.4.1 // indirect
@@ -26,7 +26,7 @@ require (
         go.uber.org/multierr v1.5.0 // indirect
         golang.org/x/crypto v0.0.0-20200510223506-06a226fb4e37
         golang.org/x/lint v0.0.0-20200302205851-738671d3881b // indirect
-        golang.org/x/net v0.0.0-20200513185701-a91f0712d120 // indirect
+        golang.org/x/net v0.0.0-20200602114024-627f9648deb9 // indirect
         golang.org/x/sync v0.0.0-20200317015054-43a5402ce75a // indirect
         golang.org/x/sys v0.0.0-20200602225109-6fdc65e7d980
         golang.org/x/tools v0.0.0-20200513201620-d5fe73897c97 // indirect
diff --git a/go.sum b/go.sum
index 5799b02..6697ba5 100644
--- a/go.sum
+++ b/go.sum
@@ -60,8 +60,8 @@ github.com/eapache/go-resiliency v1.1.0/go.mod h1:kFI+JgMyC7bLPUVY133qvEBtVayf5m
 github.com/eapache/go-xerial-snappy v0.0.0-20180814174437-776d5712da21/go.mod h1:+020luEh2TKB4/GOp8oxxtq0Daoen/Cii55CzbTV6DU=
 github.com/eapache/queue v1.1.0/go.mod h1:6eCeP0CKFpHLu8blIFXhExK/dRa7WDZfr6jVFPTqq+I=
 github.com/edsrzf/mmap-go v1.0.0/go.mod h1:YO35OhQPt3KJa3ryjFM5Bs14WD66h8eGKpfaBNrHW5M=
-github.com/ema/qdisc v0.0.0-20190904071900-b82c76788043 h1:I3hLsM87FSASssIrIOGwJCio31dvLkvpYDKn2+r31ec=
-github.com/ema/qdisc v0.0.0-20190904071900-b82c76788043/go.mod h1:ix4kG2zvdUd8kEKSW0ZTr1XLks0epFpI4j745DXxlNE=
+github.com/ema/qdisc v0.0.0-20200603082823-62d0308e3e00 h1:0GHzegkDz/zSrt+Zph1OueNImPdUxoToypnkhhRYTjI=
+github.com/ema/qdisc v0.0.0-20200603082823-62d0308e3e00/go.mod h1:ix4kG2zvdUd8kEKSW0ZTr1XLks0epFpI4j745DXxlNE=
 github.com/envoyproxy/go-control-plane v0.6.9/go.mod h1:SBwIajubJHhxtWwsL9s8ss4safvEdbitLhGGK48rN6g=
 github.com/envoyproxy/go-control-plane v0.9.1-0.20191026205805-5f8ba28d4473/go.mod h1:YTl/9mNaCwkRvm6d1a2C3ymFceY/DCBVvsKhRF0iEA4=
 github.com/envoyproxy/protoc-gen-validate v0.1.0/go.mod h1:iSmxcyjqTsJpI2R4NaDN7+kN2VEUnK/pcBlmesArF7c=
@@ -394,8 +394,8 @@ golang.org/x/net v0.0.0-20191007182048-72f939374954/go.mod h1:z5CRVTTTmAJ677TzLL
 golang.org/x/net v0.0.0-20200202094626-16171245cfb2 h1:CCH4IOTTfewWjGOlSp+zGcjutRKlBEZQ6wTn8ozI/nI=
 golang.org/x/net v0.0.0-20200202094626-16171245cfb2/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
 golang.org/x/net v0.0.0-20200226121028-0de0cce0169b/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
-golang.org/x/net v0.0.0-20200513185701-a91f0712d120 h1:EZ3cVSzKOlJxAd8e8YAJ7no8nNypTxexh/YE/xW3ZEY=
-golang.org/x/net v0.0.0-20200513185701-a91f0712d120/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
+golang.org/x/net v0.0.0-20200602114024-627f9648deb9 h1:pNX+40auqi2JqRfOP1akLGtYcn15TUbkhwuCO3foqqM=
+golang.org/x/net v0.0.0-20200602114024-627f9648deb9/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
 golang.org/x/oauth2 v0.0.0-20190226205417-e64efc72b421/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
diff --git a/vendor/github.com/ema/qdisc/get.go b/vendor/github.com/ema/qdisc/get.go
index 9191594..c83024c 100644
--- a/vendor/github.com/ema/qdisc/get.go
+++ b/vendor/github.com/ema/qdisc/get.go
@@ -86,6 +86,8 @@ type QdiscInfo struct {
         GcFlows     uint64
         Throttled   uint64
         FlowsPlimit uint64
+        Qlen        uint32
+        Backlog     uint32
 }
 
 func parseTCAStats(attr netlink.Attribute) TC_Stats {
@@ -237,6 +239,8 @@ func parseMessage(msg netlink.Message) (QdiscInfo, error) {
                         // requeues only available in TCA_STATS2, not in TCA_STATS
                         m.Requeues = s2.Requeues
                         m.Overlimits = s2.Overlimits
+                        m.Qlen = s2.Qlen
+                        m.Backlog = s2.Backlog
                 case TCA_STATS:
                         // Legacy
                         s = parseTCAStats(attr)
@@ -244,6 +248,8 @@ func parseMessage(msg netlink.Message) (QdiscInfo, error) {
                         m.Packets = s.Packets
                         m.Drops = s.Drops
                         m.Overlimits = s.Overlimits
+                        m.Qlen = s.Qlen
+                        m.Backlog = s.Backlog
                 default:
                         // TODO: TCA_OPTIONS and TCA_XSTATS
                 }
diff --git a/vendor/golang.org/x/net/http2/transport.go b/vendor/golang.org/x/net/http2/transport.go
index 54acc1e..76a92e0 100644
--- a/vendor/golang.org/x/net/http2/transport.go
+++ b/vendor/golang.org/x/net/http2/transport.go
@@ -108,6 +108,19 @@ type Transport struct {
         // waiting for their turn.
         StrictMaxConcurrentStreams bool
 
+        // ReadIdleTimeout is the timeout after which a health check using ping
+        // frame will be carried out if no frame is received on the connection.
+        // Note that a ping response will is considered a received frame, so if
+        // there is no other traffic on the connection, the health check will
+        // be performed every ReadIdleTimeout interval.
+        // If zero, no health check is performed.
+        ReadIdleTimeout time.Duration
+
+        // PingTimeout is the timeout after which the connection will be closed
+        // if a response to Ping is not received.
+        // Defaults to 15s.
+        PingTimeout time.Duration
+
         // t1, if non-nil, is the standard library Transport using
         // this transport. Its settings are used (but not its
         // RoundTrip method, etc).
@@ -131,6 +144,14 @@ func (t *Transport) disableCompression() bool {
         return t.DisableCompression || (t.t1 != nil && t.t1.DisableCompression)
 }
 
+func (t *Transport) pingTimeout() time.Duration {
+        if t.PingTimeout == 0 {
+                return 15 * time.Second
+        }
+        return t.PingTimeout
+
+}
+
 // ConfigureTransport configures a net/http HTTP/1 Transport to use HTTP/2.
 // It returns an error if t1 has already been HTTP/2-enabled.
 func ConfigureTransport(t1 *http.Transport) error {
@@ -675,6 +696,20 @@ func (t *Transport) newClientConn(c net.Conn, singleUse bool) (*ClientConn, erro
         return cc, nil
 }
 
+func (cc *ClientConn) healthCheck() {
+        pingTimeout := cc.t.pingTimeout()
+        // We don't need to periodically ping in the health check, because the readLoop of ClientConn will
+        // trigger the healthCheck again if there is no frame received.
+        ctx, cancel := context.WithTimeout(context.Background(), pingTimeout)
+        defer cancel()
+        err := cc.Ping(ctx)
+        if err != nil {
+                cc.closeForLostPing()
+                cc.t.connPool().MarkDead(cc)
+                return
+        }
+}
+
 func (cc *ClientConn) setGoAway(f *GoAwayFrame) {
         cc.mu.Lock()
         defer cc.mu.Unlock()
@@ -846,14 +881,12 @@ func (cc *ClientConn) sendGoAway() error {
         return nil
 }
 
-// Close closes the client connection immediately.
-//
-// In-flight requests are interrupted. For a graceful shutdown, use Shutdown instead.
-func (cc *ClientConn) Close() error {
+// closes the client connection immediately. In-flight requests are interrupted.
+// err is sent to streams.
+func (cc *ClientConn) closeForError(err error) error {
         cc.mu.Lock()
         defer cc.cond.Broadcast()
         defer cc.mu.Unlock()
-        err := errors.New("http2: client connection force closed via ClientConn.Close")
         for id, cs := range cc.streams {
                 select {
                 case cs.resc <- resAndError{err: err}:
@@ -866,6 +899,20 @@ func (cc *ClientConn) Close() error {
         return cc.tconn.Close()
 }
 
+// Close closes the client connection immediately.
+//
+// In-flight requests are interrupted. For a graceful shutdown, use Shutdown instead.
+func (cc *ClientConn) Close() error {
+        err := errors.New("http2: client connection force closed via ClientConn.Close")
+        return cc.closeForError(err)
+}
+
+// closes the client connection immediately. In-flight requests are interrupted.
+func (cc *ClientConn) closeForLostPing() error {
+        err := errors.New("http2: client connection lost")
+        return cc.closeForError(err)
+}
+
 const maxAllocFrameSize = 512 << 10
 
 // frameBuffer returns a scratch buffer suitable for writing DATA frames.
@@ -1737,8 +1784,17 @@ func (rl *clientConnReadLoop) run() error {
         rl.closeWhenIdle = cc.t.disableKeepAlives() || cc.singleUse
         gotReply := false // ever saw a HEADERS reply
         gotSettings := false
+        readIdleTimeout := cc.t.ReadIdleTimeout
+        var t *time.Timer
+        if readIdleTimeout != 0 {
+                t = time.AfterFunc(readIdleTimeout, cc.healthCheck)
+                defer t.Stop()
+        }
         for {
                 f, err := cc.fr.ReadFrame()
+                if t != nil {
+                        t.Reset(readIdleTimeout)
+                }
                 if err != nil {
                         cc.vlogf("http2: Transport readFrame error on conn %p: (%T) %v", cc, err, err)
                 }
diff --git a/vendor/modules.txt b/vendor/modules.txt
index 7d451f0..983f45d 100644
--- a/vendor/modules.txt
+++ b/vendor/modules.txt
@@ -14,7 +14,7 @@ github.com/cespare/xxhash/v2
 # github.com/coreos/go-systemd v0.0.0-20191104093116-d3cd4ed1dbcf
 ## explicit
 github.com/coreos/go-systemd/dbus
-# github.com/ema/qdisc v0.0.0-20190904071900-b82c76788043
+# github.com/ema/qdisc v0.0.0-20200603082823-62d0308e3e00
 ## explicit
 github.com/ema/qdisc
 # github.com/go-kit/kit v0.10.0
@@ -108,7 +108,7 @@ golang.org/x/crypto/bcrypt
 golang.org/x/crypto/blowfish
 # golang.org/x/lint v0.0.0-20200302205851-738671d3881b
 ## explicit
-# golang.org/x/net v0.0.0-20200513185701-a91f0712d120
+# golang.org/x/net v0.0.0-20200602114024-627f9648deb9
 ## explicit
 golang.org/x/net/bpf
 golang.org/x/net/http/httpguts