Alright so I've done some testing to try and track this down: I have one slice on the FV that just has ports on 130.127.39.131 from vlan (1750 and vlan 1744) I have one slice on the FV that just has ports on 130.127.39.131 on vlan1750 and another slice that only has ports on vlan1744
I also have an additional vlan on 130.127.39.131 that points directly to a controller. I've now added another switch (130.127.39.132 - pronto3290 running latest indigo) pointing at this flowvisor which I have a slice that just includes some ports on this switch. What I find is that anything that talks to the flowvisor sees these messages at the same time. Running netstat at the machine that the flowvisor is running on show thats both .39.132 and 39.131 all seem to get disconnected at the same time. And In nox I see the messages print that reflect that at the same time. The controller that is running directly from 130.127.39.131 on its own vlan never disconnects. (All of these slices are pointing to the same IP and I'm running nox's switch module on different ports for each slice). while [ 1 -eq 1 ] ; do netstat -a | grep "130.127.39.13[12]" | grep -v ESTAB ; done tcp6 0 0 expedient.clemson.:6633 130.127.39.132%13:46174 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%13:57123 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%13:64177 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.132%14:46174 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:57123 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:64177 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.132%16:46174 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%16:57123 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%16:64177 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.132%15:46174 TIME_WAIT tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:57123 TIME_WAIT Thanks, Aaron On Thu, Jul 14, 2011 at 9:11 PM, Aaron Rosen <[email protected]> wrote: > Hi Nick, > > At one point it time when the system that fv was on was really bogged > down I was able to do show openflow 1750 and it said backoff, idle, > connecting, backoff but now that I have restarted flowvisor I can't > seem to catch it. When this message prints from nox I quickly hit > entry on show openflow 1750 but I still says active. > > The switch in question here is 130.127.39.131 > > > At the machine running FV I have this going: > > while [ 1 -eq 1 ] ; do netstat -a | grep "39.131" | grep ESTAB -v ; done > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%13:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%13:62919 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:62919 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:62919 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:62919 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%14:62919 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:35160 TIME_WAIT > tcp6 0 0 expedient.clemson.:6633 130.127.39.131%15:62919 TIME_WAIT > > It seems the TCP connect gets to the TIME_WAIT state so the connection > between the switch (130.127.39.131) actually closed. > > (netstat seems to hang for a while so I didn't catch anything else. > Also everything I pasted above is what came out of netstat after the > error occurred and after that flashed it was able to go to ESTABLISHED > again. > > > P.S: Any idea why its using tcp6 sockets? It has a ipv4 address. > > Thanks, > > Aaron > > On Thu, Jul 14, 2011 at 8:43 PM, Nicholas Bastin <[email protected]> wrote: >> On Thu, Jul 14, 2011 at 19:37, Aaron Rosen <[email protected]> wrote: >>> It seems like the switch was actually connecting and disconnecting to >>> flowvisor. >> >> Is it really connecting/disconnecting from FV, or just NOX? >> >> -- >> Nick >> > > > > -- > Aaron O. Rosen > Masters Student - Network Communication > 306B Fluor Daniel > -- Aaron O. Rosen Masters Student - Network Communication 306B Fluor Daniel _______________________________________________ openflow-discuss mailing list [email protected] https://mailman.stanford.edu/mailman/listinfo/openflow-discuss
