Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[mininet-discuss] mininet simulates many switches

Bob Lantz rlantz at cs.stanford.edu
Fri Jan 16 18:26:38 PST 2015


p.s. Another thing you might want to do is to make sure that NetworkManager and IPv6 are both disabled (e.g. GRUB_CMDLINE_LINUX_DEFAULT="ipv6.disable=1" ), and also check your CPU load and memory usage.

> On Jan 16, 2015, at 6:22 PM, Bob Lantz <rlantz at cs.stanford.edu> wrote:
> 
> Hmm, maybe something is odd with your environment.
> 
> 1. If you run with -v debug, do you see any errors?
> 
> 2. When you look at the switches with ‘ovs-vsctl show’ does everything look good?
> 
> 3. If you try the latest Mininet VM image, do you see the same behavior?
> 
> If the VM image works OK, you might want to compare it with your test environment to see if any differences could be causing problems.
> 
> If the VM image doesn’t work, then I’m not sure…  perhaps you might try on different machines, for example a server with a fast CPU and lots of memory, and also try giving more resources to the VM.
> 
> In the worst case, well, you could try downsizing your experiment to a smaller number of switches and hosts, or running across multiple servers.
> 
> Good luck.
> 
> On Jan 16, 2015, at 5:39 PM, rui liu <lr1990311 at sina.com> wrote:
>> 
>> I am just doing an experiment, so connected many switches.
>> There are the result of reference controller in my environment:
>> 
>> UServer102:~/scripts$ sudo mn --topo linear,300 --controller ref -v output
>> mininet> h1 ping h237
>> PING 10.0.0.237 (10.0.0.237) 56(84) bytes of data.
>> From 10.0.0.1 icmp_seq=1 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=2 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=3 Destination Host Unreachable
>> 64 bytes from 10.0.0.237: icmp_req=4 ttl=64 time=3333 ms
>> 64 bytes from 10.0.0.237: icmp_req=5 ttl=64 time=2326 ms
>> 64 bytes from 10.0.0.237: icmp_req=6 ttl=64 time=1318 ms
>> 64 bytes from 10.0.0.237: icmp_req=7 ttl=64 time=310 ms
>> 64 bytes from 10.0.0.237: icmp_req=8 ttl=64 time=35.1 ms
>> ^C
>> Interrupt
>> mininet> h1 ping h238
>> PING 10.0.0.238 (10.0.0.238) 56(84) bytes of data.
>> From 10.0.0.1 icmp_seq=1 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=2 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=3 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=4 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=5 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=6 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=7 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=8 Destination Host Unreachable
>> From 10.0.0.1 icmp_seq=9 Destination Host Unreachable
>> ^C
>> Interrupt
>> mininet> h237 ping h238
>> PING 10.0.0.238 (10.0.0.238) 56(84) bytes of data.
>> From 10.0.0.237 icmp_seq=1 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=2 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=3 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=4 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=5 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=6 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=7 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=8 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=9 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=10 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=11 Destination Host Unreachable
>> From 10.0.0.237 icmp_seq=12 Destination Host Unreachable
>> ^C
>> Interrupt
>> 
>> In the result, first three packet of "h1 ping h237" is Host unreachable.
>> 
>> 2015-01-17 5:49 GMT+08:00 Bob Lantz <rlantz at cs.stanford.edu>:
>> 1. Please don’t drop the list.
>> 
>> 2. Not that I’m aware of. I would suspect that “host unreachable” could be caused by a timeout or dropped packet. Can you confirm that the reference controller fails on your system?
>> 
>> 3. Note that this is a somewhat unusual configuration - it’s rare in a real network for a packet to traverse 300 switches.
>> 
>>> On Jan 15, 2015, at 10:39 PM, rui liu <lr1990311 at sina.com> wrote:
>>> 
>>> Thank you for replying me.
>>> Do you  have other controller like Floodlight,in my result,it was "Host Unreachable" ,may be not timeout.
>>> And I am also use the same command like you in the mininet. The result is h1 can't ping h300. So i wonder if there are some configuration different with mininet?or invlved with os parameters?
>>> 
>>> 2015-01-16 11:45 GMT+08:00 Bob Lantz <rlantz at cs.stanford.edu>:
>>> 
>>> I tried it with the reference controller; note this reactive controller relies on packet_ins/outs at every hop, so it should not be surprising that the first few replies can be delayed (possibly resulting in ping timeout) or reordered (later packets can take the fast path while earlier packets are sent via packet_out; moreover, OVS might not push flows immediately into the kernel flow cache):
>>> 
>>> openflow at ubuntu13:~/mininet$ sudo mn --topo linear,300 --controller ref -v output
>>> mininet> h1 ping h300
>>> PING 10.0.1.44 (10.0.1.44) 56(84) bytes of data.
>>> 64 bytes from 10.0.1.44: icmp_seq=1 ttl=64 time=6673 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=2 ttl=64 time=5664 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=4 ttl=64 time=3665 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=6 ttl=64 time=1665 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=7 ttl=64 time=664 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=5 ttl=64 time=2666 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=3 ttl=64 time=4666 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=8 ttl=64 time=16.6 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=9 ttl=64 time=6.92 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=10 ttl=64 time=1.29 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=11 ttl=64 time=14.5 ms
>>> 64 bytes from 10.0.1.44: icmp_seq=12 ttl=64 time=14.5 ms
>>> ^C
>>> --- 10.0.1.44 ping statistics ---
>>> 12 packets transmitted, 12 received, 0% packet loss, time 11022ms
>>> rtt min/avg/max/mdev = 1.293/2143.288/6673.211/2363.386 ms, pipe 7
>>> 
>>> 
>>> > On Jan 15, 2015, at 5:27 PM, rui liu <lr1990311 at sina.com> wrote:
>>> >
>>> > Hello everyone, I am having a problem with mininet. I used mn to create a topo in type of linear with 300 switches, the command are:
>>> >
>>> >  sudo mn --topo=linear,300 --controller "remote,ip=192.168.10.101"
>>> >
>>> > mininet connect with a remote controller floodlight.The result is hosts connect on s1-s237 can ping each other,but hosts connect on s237-s300 can't ping each other,the result are:
>>> >
>>> > mininet> h1 ping h100
>>> > PING 10.0.0.100 (10.0.0.100) 56(84) bytes of data.
>>> > 64 bytes from 10.0.0.100: icmp_req=1 ttl=64 time=1516 ms
>>> > 64 bytes from 10.0.0.100: icmp_req=2 ttl=64 time=516 ms
>>> >
>>> > mininet> h1 ping h237
>>> > PING 10.0.0.237 (10.0.0.237) 56(84) bytes of data.
>>> > 64 bytes from 10.0.0.237: icmp_req=1 ttl=64 time=2483 ms
>>> > 64 bytes from 10.0.0.237: icmp_req=2 ttl=64 time=1484 ms
>>> >
>>> > mininet> h237 ping h238
>>> > PING 10.0.0.238 (10.0.0.238) 56(84) bytes of data.
>>> > From 10.0.0.237 icmp_seq=1 Destination Host Unreachable
>>> > From 10.0.0.237 icmp_seq=2 Destination Host Unreachable
>>> >
>>> > mininet> h238 ping h300
>>> > PING 10.0.1.44 (10.0.1.44) 56(84) bytes of data.
>>> > From 10.0.0.238 icmp_seq=1 Destination Host Unreachable
>>> > From 10.0.0.238 icmp_seq=2 Destination Host Unreachable
>>> >
>>> > I parsed the Packet_in in the floodlight,and found there are no flow_request beyond s237(beyond S237 are only LLDPs and device attachpoints packets).So I wonder if there are some parameters mininet used involved with OS?
>>> >
>>> > I am also using some other SDN controllers like POX, ONOS, Opendaylight. The results are similiar, but the partition switch is different,may be s137,s50 or else.
>>> >
>>> > Do you have the same problem?
>>> > _______________________________________________
>>> > mininet-discuss mailing list
>>> > mininet-discuss at lists.stanford.edu
>>> > https://mailman.stanford.edu/mailman/listinfo/mininet-discuss
>>> 
>>> 
>>> 
>>> 
>> 
>> 
> 



More information about the mininet-discuss mailing list