hi, If you look at the log file there are two entries
11:28:26 root : INFO : *** TEST RUN START: Thu May 3 11:28:26 2012 runTest (basic.DataPlaneOnly) ... 11:28:26 basic : INFO : ** START DataPlaneOnly CASE runTest (basic.DataPlaneOnly) 11:28:26 dp-eth3 : INFO : Could not open socket ERROR runTest (basic.DescStatsGet) ... 11:28:26 basic : INFO : ** START TEST CASE runTest (basic.DescStatsGet) 11:28:26 controller: INFO : Create/listen at 127.0.0.1:6634 11:28:26 controller: INFO : Waiting for switch connection Controller startup failed (no switch addr); exiting ERROR Probably this started the controller as "create/listen at 127.0.0.1:6634" must have passed and it says waiting for switch connection.. Again you have to make sure that the switch is now trying to connect to 6634 rather than 6633. It is followed by another message "Controller startup failed.." probably something wrong with the requirements to the controller.. but does not say whether it did a clean exit by quitting everything from 6634. The next test starts and tries to connect to 6634 and probably it is still occupied.. It is nothing to do with your controller not running before starting this.. but after starting this the controller probably is started and stopped for everytest. Couple of suggestions: a) Also please use these parameters in addition .. sudo ./oft --host=127.0.0.1 --port=6633 --verbose --test-spec=Echo where host is the IP of the interface on the test server to which the switch can connect b) One test to verify is that you can run just one test.. but giving specific parameter.. like ./oft --test-spec=TableStatsGet and see whether it succeeds. or --test-spec=FillTableExact you get the list of test modules by doing ./oft --list. You can see if any of them succeeds. You can also specifically look at http://forums.openflowhub.org/showthread.php?86-Newbie-question-Packein-Test-failingand http://forums.openflowhub.org/showthread.php?89-Not-able-to-execute-OFTestfor some information. On Thu, May 3, 2012 at 3:00 PM, Xuan-Nam Nguyen <xuan-nam.ngu...@inria.fr>wrote: > Thanks for reply. > > a. I tried but it does not work > b. I make sure that nothing is running on port 6633 (by netstat -ltnup) > and no other oft instance (by ps -aux | grep oft) > Even when I tried with other port, I still got same error (Log file) > > Nam > > > > > On 05/03/2012 11:17 AM, kalapriya k wrote: > > hi, > > Two things: > > a) if your switch is running remotely, you have to specific to the oftest > that it is remote. This can be done with the --platform=remote. > > b)the error in the logfile seems to indicate that the port 6633 on your > machine (in which you are running oftest) is not free. probably there is a > controller already running on that port. Stop any other controller that is > running on the machine and start your oft. Oft will start a controller on > that port and the switch will connect to it. > > hope it helps. > > On Thu, May 3, 2012 at 2:37 PM, Xuan-Nam Nguyen > <xuan-nam.ngu...@inria.fr>wrote: > >> Thanks for reply. >> >> 1. I attach in this mail the startUp log of pronto 3290 switch, may be it >> is useful. >> >> 2. I check out the newest source code of OFTest and compile it following >> the README >> I make sure that >> the OF deamon Pronto switch is running and trying to connect to >> controller (I tested with my controller and I works well) >> nothing runs on port 6633 >> >> But when I started with command : >> ./oft --test-spec=Echo >> ./oft --verbose --log-file="" >> >> I got error (see the attached file) >> Do you know what is the problem ? >> Thanks >> >> Nam >> ** >> ------------------------------ >> >> *From: *"Dan Talayco" <dan.tala...@bigswitch.com> >> *To: *"Xuan-Nam Nguyen" <xuan-nam.ngu...@inria.fr> >> *Cc: *openflow-discuss@lists.stanford.edu, "Damien Saucez" < >> damien.sau...@inria.fr> >> *Sent: *Wednesday, May 2, 2012 6:35:51 PM >> *Subject: *Re: [openflow-discuss] Pronto 3290 max flow entries ? >> >> >> Hi-- >> >> I just ran OFTest against a 3290 in our lab running the test >> FillTableExact. I got >> >> RESULT: 3840 flows reported >> >> a bit below the 4K, but about right. I'm not aware of any flow types >> that might result in this kind of limit (2K that you're seeing). >> >> I would suggest trying to run FillTableExact from OFTest (I just >> checked in a couple of bug fixes to the git repo, so make sure you get the >> latest updates). It's at https://github.com/floodlight/oftest/ >> >> Do you know where this 3290 came from? There may be some older >> versions that had different ASICs in them that had other limits. I'm not >> sure how to determine if that's the case, but someone at Pronto/Pica8 might >> know. >> >> -Dan >> >> >> >> On Wednesday, May 2, 2012 at 7:50 AM, Xuan-Nam Nguyen wrote: >> >> Hi all, >> >> As the specification on >> http://yuba.stanford.edu/foswiki/bin/view/OpenFlow/Deployment/Vendor/Pronto: >> Pronto 3290 has maximum ~ 4000 flow entries >> We have an Pronto Switch 3290 which used newest firmware (2011-03-19 >> http://www.openflowhub.org/display/Indigo/Indigo+-+Download+Page) >> But we can only insert maximum 1920 *exact match* flow entries into that >> switch. In controller (Beacon) we got error OFPET_FLOW_MOD_FAILED >> OFPFMFC_ALL_TABLES_FULL >> >> Do you know what is the reason ? >> >> Thanks >> >> NGUYEN Xuan Nam >> INRIA Sophia >> >> >> >> >> >> >> _______________________________________________ >> openflow-discuss mailing list >> openflow-discuss@lists.stanford.edu >> https://mailman.stanford.edu/mailman/listinfo/openflow-discuss >> >> Attachments: >> - Screenshot-Indigo OpenFlow Switch - Mozilla Firefox.png >> - Screenshot-Beacon - Mozilla Firefox.png >> >> >> >> >> _______________________________________________ >> openflow-discuss mailing list >> openflow-discuss@lists.stanford.edu >> https://mailman.stanford.edu/mailman/listinfo/openflow-discuss >> >> > > > -- > NGUYEN Xuan Nam >
_______________________________________________ openflow-discuss mailing list openflow-discuss@lists.stanford.edu https://mailman.stanford.edu/mailman/listinfo/openflow-discuss
