tipc: add name distributor resiliency queue

TIPC name table updates are distributed asynchronously in a cluster,
entailing a risk of certain race conditions. E.g., if two nodes
simultaneously issue conflicting (overlapping) publications, this may
not be detected until both publications have reached a third node, in
which case one of the publications will be silently dropped on that
node. Hence, we end up with an inconsistent name table.

In most cases this conflict is just a temporary race, e.g., one
node is issuing a publication under the assumption that a previous,
conflicting, publication has already been withdrawn by the other node.
However, because of the (rtt related) distributed update delay, this
may not yet hold true on all nodes. The symptom of this failure is a
syslog message: "tipc: Cannot publish {%u,%u,%u}, overlap error".

In this commit we add a resiliency queue at the receiving end of
the name table distributor. When insertion of an arriving publication
fails, we retain it in this queue for a short amount of time, assuming
that another update will arrive very soon and clear the conflict. If so
happens, we insert the publication, otherwise we drop it.

The (configurable) retention value defaults to 2000 ms. Knowing from
experience that the situation described above is extremely rare, there
is no risk that the queue will accumulate any large number of items.

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

1 files changed, 16 insertions, 0 deletions

diff --git a/Documentation/sysctl/net.txt b/Documentation/sysctl/net.txt
index 9a0319a82470..04892b821157 100644
--- a/Documentation/sysctl/net.txt
+++ b/Documentation/sysctl/net.txt
@@ -241,6 +241,9 @@ address of the router (or Connected) for internal networks.
 6. TIPC
 -------------------------------------------------------
 
+tipc_rmem
+----------
+
 The TIPC protocol now has a tunable for the receive memory, similar to the
 tcp_rmem - i.e. a vector of 3 INTEGERs: (min, default, max)
 
@@ -252,3 +255,16 @@ The max value is set to CONN_OVERLOAD_LIMIT, and the default and min values
 are scaled (shifted) versions of that same value.  Note that the min value
 is not at this point in time used in any meaningful way, but the triplet is
 preserved in order to be consistent with things like tcp_rmem.
+
+named_timeout
+--------------
+
+TIPC name table updates are distributed asynchronously in a cluster, without
+any form of transaction handling. This means that different race scenarios are
+possible. One such is that a name withdrawal sent out by one node and received
+by another node may arrive after a second, overlapping name publication already
+has been accepted from a third node, although the conflicting updates
+originally may have been issued in the correct sequential order.
+If named_timeout is nonzero, failed topology updates will be placed on a defer
+queue until another event arrives that clears the error, or until the timeout
+expires. Value is in milliseconds.