tipc: avoid stale link after aborted failover
authorJon Paul Maloy <jon.maloy@ericsson.com>
Tue, 3 Feb 2015 13:59:18 +0000 (08:59 -0500)
committerDavid S. Miller <davem@davemloft.net>
Thu, 5 Feb 2015 00:09:31 +0000 (16:09 -0800)
During link failover it may happen that the remaining link goes
down while it is still in the process of taking over traffic
from a previously failed link. When this happens, we currently
abort the failover procedure and reset the first failed link to
non-failover mode, so that it will be ready to re-establish
contact with its peer when it comes available.

However, if the first link goes down because its bearer was manually
disabled, it is not enough to reset it; it must also be deleted;
which is supposed to happen when the failover procedure is finished.
Otherwise it will remain a zombie link: attached to the owner node
structure, in mode LINK_STOPPED, and permanently blocking any re-
establishing of the link to the peer via the interface in question.

We fix this by amending the failover abort procedure. Apart from
resetting the link to non-failover state, we test if the link is
also in LINK_STOPPED mode. If so, we delete it, using the conditional
tipc_link_delete() function introduced in the previous commit.

Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-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>
net/tipc/link.h
net/tipc/node.c

index f06b779c9f7533b868e6b9d4ad53370b6b928239..3e3432b3044e01c867feb2d0e04af9b102531082 100644 (file)
@@ -202,6 +202,7 @@ struct tipc_port;
 struct tipc_link *tipc_link_create(struct tipc_node *n_ptr,
                              struct tipc_bearer *b_ptr,
                              const struct tipc_media_addr *media_addr);
+void tipc_link_delete(struct tipc_link *link);
 void tipc_link_delete_list(struct net *net, unsigned int bearer_id,
                           bool shutting_down);
 void tipc_link_failover_send_queue(struct tipc_link *l_ptr);
index ee5d33cfcf8032ffc40b3c029471ecf75b34015a..d4cb8c127063d73a963e62fe1cd230a1ef3f348a 100644 (file)
@@ -406,6 +406,10 @@ static void node_lost_contact(struct tipc_node *n_ptr)
                l_ptr->reset_checkpoint = l_ptr->next_in_no;
                l_ptr->exp_msg_count = 0;
                tipc_link_reset_fragments(l_ptr);
+
+               /* Link marked for deletion after failover? => do it now */
+               if (l_ptr->flags & LINK_STOPPED)
+                       tipc_link_delete(l_ptr);
        }
 
        n_ptr->action_flags &= ~TIPC_WAIT_OWN_LINKS_DOWN;