tipc: add neighbor monitoring framework

TIPC based clusters are by default set up with full-mesh link
connectivity between all nodes. Those links are expected to provide
a short failure detection time, by default set to 1500 ms. Because
of this, the background load for neighbor monitoring in an N-node
cluster increases with a factor N on each node, while the overall
monitoring traffic through the network infrastructure increases at
a ~(N * (N - 1)) rate. Experience has shown that such clusters don't
scale well beyond ~100 nodes unless we significantly increase failure
discovery tolerance.

This commit introduces a framework and an algorithm that drastically
reduces this background load, while basically maintaining the original
failure detection times across the whole cluster. Using this algorithm,
background load will now grow at a rate of ~(2 * sqrt(N)) per node, and
at ~(2 * N * sqrt(N)) in traffic overhead. As an example, each node will
now have to actively monitor 38 neighbors in a 400-node cluster, instead
of as before 399.

This "Overlapping Ring Supervision Algorithm" is completely distributed
and employs no centralized or coordinated state. It goes as follows:

- Each node makes up a linearly ascending, circular list of all its N
  known neighbors, based on their TIPC node identity. This algorithm
  must be the same on all nodes.

- The node then selects the next M = sqrt(N) - 1 nodes downstream from
  itself in the list, and chooses to actively monitor those. This is
  called its "local monitoring domain".

- It creates a domain record describing the monitoring domain, and
  piggy-backs this in the data area of all neighbor monitoring messages
  (LINK_PROTOCOL/STATE) leaving that node. This means that all nodes in
  the cluster eventually (default within 400 ms) will learn about
  its monitoring domain.

- Whenever a node discovers a change in its local domain, e.g., a node
  has been added or has gone down, it creates and sends out a new
  version of its node record to inform all neighbors about the change.

- A node receiving a domain record from anybody outside its local domain
  matches this against its own list (which may not look the same), and
  chooses to not actively monitor those members of the received domain
  record that are also present in its own list. Instead, it relies on
  indications from the direct monitoring nodes if an indirectly
  monitored node has gone up or down. If a node is indicated lost, the
  receiving node temporarily activates its own direct monitoring towards
  that node in order to confirm, or not, that it is actually gone.

- Since each node is actively monitoring sqrt(N) downstream neighbors,
  each node is also actively monitored by the same number of upstream
  neighbors. This means that all non-direct monitoring nodes normally
  will receive sqrt(N) indications that a node is gone.

- A major drawback with ring monitoring is how it handles failures that
  cause massive network partitionings. If both a lost node and all its
  direct monitoring neighbors are inside the lost partition, the nodes in
  the remaining partition will never receive indications about the loss.
  To overcome this, each node also chooses to actively monitor some
  nodes outside its local domain. Those nodes are called remote domain
  "heads", and are selected in such a way that no node in the cluster
  will be more than two direct monitoring hops away. Because of this,
  each node, apart from monitoring the member of its local domain, will
  also typically monitor sqrt(N) remote head nodes.

- As an optimization, local list status, domain status and domain
  records are marked with a generation number. This saves senders from
  unnecessarily conveying  unaltered domain records, and receivers from
  performing unneeded re-adaptations of their node monitoring list, such
  as re-assigning domain heads.

- As a measure of caution we have added the possibility to disable the
  new algorithm through configuration. We do this by keeping a threshold
  value for the cluster size; a cluster that grows beyond this value
  will switch from full-mesh to ring monitoring, and vice versa when
  it shrinks below the value. This means that if the threshold is set to
  a value larger than any anticipated cluster size (default size is 32)
  the new algorithm is effectively disabled. A patch set for altering the
  threshold value and for listing the table contents will follow shortly.

- This change is fully backwards compatible.

Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 37 insertions, 12 deletions

diff --git a/net/tipc/link.c b/net/tipc/link.c
index a904ccd5a93a..03f8bdf70d8f 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -42,6 +42,7 @@
 #include "name_distr.h"
 #include "discover.h"
 #include "netlink.h"
+#include "monitor.h"
 
 #include <linux/pkt_sched.h>
 
@@ -95,6 +96,7 @@ struct tipc_stats {
  * @pmsg: convenience pointer to "proto_msg" field
  * @priority: current link priority
  * @net_plane: current link network plane ('A' through 'H')
+ * @mon_state: cookie with information needed by link monitor
  * @backlog_limit: backlog queue congestion thresholds (indexed by importance)
  * @exp_msg_count: # of tunnelled messages expected during link changeover
  * @reset_rcv_checkpt: seq # of last acknowledged message at time of link reset
@@ -138,6 +140,7 @@ struct tipc_link {
 	char if_name[TIPC_MAX_IF_NAME];
 	u32 priority;
 	char net_plane;
+	struct tipc_mon_state mon_state;
 	u16 rst_cnt;
 
 	/* Failover/synch */
@@ -708,18 +711,25 @@ int tipc_link_timeout(struct tipc_link *l, struct sk_buff_head *xmitq)
 	bool setup = false;
 	u16 bc_snt = l->bc_sndlink->snd_nxt - 1;
 	u16 bc_acked = l->bc_rcvlink->acked;
-
-	link_profile_stats(l);
+	struct tipc_mon_state *mstate = &l->mon_state;
 
 	switch (l->state) {
 	case LINK_ESTABLISHED:
 	case LINK_SYNCHING:
-		if (l->silent_intv_cnt > l->abort_limit)
-			return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
 		mtyp = STATE_MSG;
+		link_profile_stats(l);
+		tipc_mon_get_state(l->net, l->addr, mstate, l->bearer_id);
+		if (mstate->reset || (l->silent_intv_cnt > l->abort_limit))
+			return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
 		state = bc_acked != bc_snt;
-		probe = l->silent_intv_cnt;
-		l->silent_intv_cnt++;
+		state |= l->bc_rcvlink->rcv_unacked;
+		state |= l->rcv_unacked;
+		state |= !skb_queue_empty(&l->transmq);
+		state |= !skb_queue_empty(&l->deferdq);
+		probe = mstate->probing;
+		probe |= l->silent_intv_cnt;
+		if (probe || mstate->monitoring)
+			l->silent_intv_cnt++;
 		break;
 	case LINK_RESET:
 		setup = l->rst_cnt++ <= 4;
@@ -830,6 +840,7 @@ void tipc_link_reset(struct tipc_link *l)
 	l->stats.recv_info = 0;
 	l->stale_count = 0;
 	l->bc_peer_is_up = false;
+	memset(&l->mon_state, 0, sizeof(l->mon_state));
 	tipc_link_reset_stats(l);
 }
 
@@ -1238,6 +1249,9 @@ static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
 	struct tipc_msg *hdr;
 	struct sk_buff_head *dfq = &l->deferdq;
 	bool node_up = link_is_up(l->bc_rcvlink);
+	struct tipc_mon_state *mstate = &l->mon_state;
+	int dlen = 0;
+	void *data;
 
 	/* Don't send protocol message during reset or link failover */
 	if (tipc_link_is_blocked(l))
@@ -1250,12 +1264,13 @@ static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
 		rcvgap = buf_seqno(skb_peek(dfq)) - l->rcv_nxt;
 
 	skb = tipc_msg_create(LINK_PROTOCOL, mtyp, INT_H_SIZE,
-			      TIPC_MAX_IF_NAME, l->addr,
+			      tipc_max_domain_size, l->addr,
 			      tipc_own_addr(l->net), 0, 0, 0);
 	if (!skb)
 		return;
 
 	hdr = buf_msg(skb);
+	data = msg_data(hdr);
 	msg_set_session(hdr, l->session);
 	msg_set_bearer_id(hdr, l->bearer_id);
 	msg_set_net_plane(hdr, l->net_plane);
@@ -1271,14 +1286,18 @@ static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
 
 	if (mtyp == STATE_MSG) {
 		msg_set_seq_gap(hdr, rcvgap);
-		msg_set_size(hdr, INT_H_SIZE);
 		msg_set_probe(hdr, probe);
+		tipc_mon_prep(l->net, data, &dlen, mstate, l->bearer_id);
+		msg_set_size(hdr, INT_H_SIZE + dlen);
+		skb_trim(skb, INT_H_SIZE + dlen);
 		l->stats.sent_states++;
 		l->rcv_unacked = 0;
 	} else {
 		/* RESET_MSG or ACTIVATE_MSG */
 		msg_set_max_pkt(hdr, l->advertised_mtu);
-		strcpy(msg_data(hdr), l->if_name);
+		strcpy(data, l->if_name);
+		msg_set_size(hdr, INT_H_SIZE + TIPC_MAX_IF_NAME);
+		skb_trim(skb, INT_H_SIZE + TIPC_MAX_IF_NAME);
 	}
 	if (probe)
 		l->stats.sent_probes++;
@@ -1371,7 +1390,9 @@ static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
 	u16 peers_tol = msg_link_tolerance(hdr);
 	u16 peers_prio = msg_linkprio(hdr);
 	u16 rcv_nxt = l->rcv_nxt;
+	u16 dlen = msg_data_sz(hdr);
 	int mtyp = msg_type(hdr);
+	void *data;
 	char *if_name;
 	int rc = 0;
 
@@ -1381,6 +1402,10 @@ static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
 	if (tipc_own_addr(l->net) > msg_prevnode(hdr))
 		l->net_plane = msg_net_plane(hdr);
 
+	skb_linearize(skb);
+	hdr = buf_msg(skb);
+	data = msg_data(hdr);
+
 	switch (mtyp) {
 	case RESET_MSG:
 
@@ -1391,8 +1416,6 @@ static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
 		/* fall thru' */
 
 	case ACTIVATE_MSG:
-		skb_linearize(skb);
-		hdr = buf_msg(skb);
 
 		/* Complete own link name with peer's interface name */
 		if_name =  strrchr(l->name, ':') + 1;
@@ -1400,7 +1423,7 @@ static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
 			break;
 		if (msg_data_sz(hdr) < TIPC_MAX_IF_NAME)
 			break;
-		strncpy(if_name, msg_data(hdr),	TIPC_MAX_IF_NAME);
+		strncpy(if_name, data, TIPC_MAX_IF_NAME);
 
 		/* Update own tolerance if peer indicates a non-zero value */
 		if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL))
@@ -1448,6 +1471,8 @@ static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
 				rc = TIPC_LINK_UP_EVT;
 			break;
 		}
+		tipc_mon_rcv(l->net, data, dlen, l->addr,
+			     &l->mon_state, l->bearer_id);
 
 		/* Send NACK if peer has sent pkts we haven't received yet */
 		if (more(peers_snd_nxt, rcv_nxt) && !tipc_link_is_synching(l))