ovs-vsctl commands like 'ovs-vsctl list Interface p1' use the
'monitor' RPC method, which causes ovsdb sending updates to
the command session when changes are committed to the monitored
table. Since ovs-vsctl commands are short-lived, there is chance
that ovs-vsctl terminates the connection to ovsdb right before
ovsdb sends the update. This race will cause the following
warning entries in ovsdb-server log:
|jsonrpc|WARN|unix: receive error: Connection reset by peer
|reconnect|WARN|unix: connection dropped (Connection reset by peer)
|jsonrpc|WARN|unix: send error: Broken pipe
|reconnect|WARN|unix: connection dropped (Broken pipe)
The bfd:flap_count test is particularly prone to this race,
since the test aligns the statistics updates (every 5 seconds)
with the invocation of ovs-vsctl commands.
In the short term, this commit fixes the intermittent failure
by disabling the ovs-vswitchd statistics updates using a huge
update interval.
In the long run, we will research on making ovsdb not send
further updates to sessions like ovs-vsctl.
Signed-off-by: Alex Wang <al...@nicira.com>
---
tests/bfd.at | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/bfd.at b/tests/bfd.at
index 609c0ad..1ccdc1b 100644
--- a/tests/bfd.at
+++ b/tests/bfd.at
@@ -684,6 +684,7 @@ OVS_VSWITCHD_START([add-br br1 -- \
set Interface p1 bfd:enable=true bfd:min_tx=100
bfd:min_rx=100])
ovs-appctl time/stop
+AT_CHECK([ovs-vsctl set Open_vSwitch .
other_config:stats-update-interval=50000000])
# Part-1 wait for a while to stablize bfd.
ovs-appctl time/warp 10100 100
--
1.7.9.5
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
