The learning switch functionality included in OpenFlow will send packets
out the port they arrived on if the learned destination is from that port.
This causes problems when newer switches flood packets for destinations they
don't know and OpenFlow reinjects packets back into the network. The correct
behavior is to just drop the packet.
When a vconn reports an error, the rconn would not reliably reconnect.
In particular, if the error was reported after the call to rconn_run()
but before rconn_run_wait() was called, then the state's "run" routine
would not set min_timeout properly, leading to a potentially arbitrarily
long wait (depending on what other events were going on in) until the
state's "run" routine was called again.
The fix is to have a separate per-state "timeout" routine to compute
when the state needs to be re-entered.
This commit was tested using the following change to randomly inject
errors:
@@ -554,11 +554,16 @@
static int
try_send(struct rconn *rc)
{
int retval = 0;
struct buffer *next = rc->txq.head->next;
- retval = vconn_send(rc->vconn, rc->txq.head);
+ if (!random_range(1000)) {
+ fprintf(stderr, "injecting ECONNRESET\n");
+ retval = ECONNRESET;
+ } else {
+ retval = vconn_send(rc->vconn, rc->txq.head);
+ }
if (retval) {
if (retval != EAGAIN) {
disconnect(rc, retval);
}
return retval;
Before, the amount of time disconnected was measured relative to the
last time we connected, that is, the last time connect(2) succeeded.
Thus, if we were connected for a long time, and then disconnected,
we would immediately go into fail-open mode.
This change make the disconnected time relative to the last OpenFlow
message received. Thus, if we are connected for a long time and
receive plenty of packets, and then disconnect, there will still be
an opportunity to reconnect before failing open.
This was in there because at one time it seemed desirable to configure
from the postinst, but that idea was abandoned. Now Debconf seems to
be screwing something up somehow, so we might as well not run it at all,
since it wasn't doing anything useful.
This fixes "module-assistant auto-install openflow-datapath" for
bootstrapping from a clean environment. (Use "m-a -f purge
openflow-datapath" to test that it works starting from an un-clean
environment.)
Generating them unconditionally caused a problem with the init script:
the script required there to be a certificate for the private key if
the private key existed, but we do want to allow TCP-only configurations
and always having a private key prevented that.
Also, give the user advice on how to deal with this problem when it
arises in the init script.
We use SIGALRM to limit runtime of dpctl and ofp-discover. There is no
reason that we should not clean up after it in the same way as any other
catchable signal. In particular, ofp-discover wants to restore network
device flags on timeout.
The upcoming ofp-discover program wants to print out the binding
information, and dhcp_msg_to_string() is pretty close to what it
wants. But one-field-per-line is easier for other programs to parse.
This allows us to have DHCP clients that don't actually bind or unbind
a network device's IP address. This is useful for doing controller
discovery without actually reconfiguring a network device.
There's no need to set changed unconditionally in dhclient_init(), and
it can make extra work for the users of the dhclient, by causing them
to think that something changed when it didn't.
libtool was needed when NOX was directly linking against libopenflow.
Since it doesn't do that anymore, we can drop its use (and that makes
OpenFlow configure and build slightly faster).
This keeps vlogconf, in particular, from being linked against OpenSSL,
which in turn keeps the Debian package build tools from complaining that
the openflow-common package shouldn't need to depend on OpenSSL but does
because of the spurious library link.
Ben Pfaff
Natasha Gude
Justin Pettit