ZK S | 28 Jun 17:32 2016
Picon

OF Experimenter tuorial

Hello ladies and gentlemen,
Do you have please any information, tutorial or examples on how creating and implementing an OF experimenter?
Thanks and regards, 
Meilleures salutations,
Salaheddine Zerkane
Rennes, France
<div><div dir="ltr">
<div class="gmail_default">Hello ladies and gentlemen,</div>
<div class="gmail_default">Do you have please any information, tutorial or examples on how creating and implementing an OF experimenter?</div>
<div class="gmail_default">Thanks and regards,&nbsp;</div>
<div><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr">
<div><span>Meilleures salutations,</span></div>Salaheddine Zerkane<div>Rennes, France</div>
</div></div></div></div></div>
</div></div>
Hssan Driss | 27 Jun 18:14 2016
Picon
Gravatar

Re: [floodlight-dev] Flow Rule not applied on the Switch Level

hey Ryan ,

You means when using dpctl add-flow ?
Do you know why ACL command doesn't get to block the port of that ip address ?

Thanks in advance 



On Mon, Jun 27, 2016 at 5:05 PM, Ryan Izard <rizard <at> g.clemson.edu> wrote:
You need to specify the dl_type (ethertype) as IPv4 (unless you're using IPv6, in which case it'll be IPv6).

Ryan

On Jun 27, 2016, at 9:54 AM, Hssan Driss <hssan.driss-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:

Hi everyone, 

Yes, as a matter of fact, on Floodlight ACL choosing UDP protocol is translated into 11. Is that decimal or hex ? 
There might be a confusion though ! 0x11 = 17 (dec) 
And Since I'm using SIP on default 5060 port shall I precise it ?

In addition, I tried adding the flow on the switch manually using dpctl commands as below : 

# sudo dpctl add-flow unix:/var/run/dp0.sock in_port=*,dl_vlan=*,dl_src=*,dl_dst=*,dl_type=*,nw_src=192.168.2.2,nw_dst=*,nw_proto=17,tp_src=*,tp_dst=5060,icmp_type=*,icmp_code=*,actions=

I get : 

# sudo dpctl dump-flows unix:/var/run/dp0.sock
stats_reply (xid=0x54e91f95): flags=none type=1(flow)
  cookie=0, duration_sec=1s, duration_nsec=552000000s, table_id=0, priority=32768, n_packets=0, n_bytes=0, idle_timeout=60,hard_timeout=0,actions=

And packets are blocked ! but can't get why the flow rule doesn't exactly match the rule added !

Thanks in advance for your help ! 

Kind Regards, 
Hssan 


On Fri, Jun 24, 2016 at 4:25 PM, Hssan Driss <hssan.driss <at> gmail.com> wrote:
Hi everyone,

I'm working on SDN lab. When intend to block SIP Client IP. We are pushing the blocking order based on the Src-ip address to Floodlight controller through REST API. Which is translated on a switch level :

# dpctl dump-flows unix:/var/run/dp0.sock
stats_reply (xid=0x79853a73): flags=none type=1(flow)
  cookie=45035999707555269, duration_sec=9s, duration_nsec=451000000s, table_id=0, priority=29999, n_packets=0, n_bytes=0, idle_timeout=0,hard_timeout=0,ip,nw_src=192.168.2.2,nw_proto=11,tp_src=0,tp_dst=0,actions=
  cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=514, n_bytes=231936, idle_timeout=5,hard_timeout=0,in_port=4,dl_src=00:21:9b:88:04:66,dl_dst=18:03:73:b5:78:4d,actions=output:3
  cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=685, n_bytes=383282, idle_timeout=5,hard_timeout=0,in_port=3,dl_src=18:03:73:b5:78:4d,dl_dst=00:21:9b:88:04:66,actions=output:4

SIP Client is on 192.168.2.2

The flow rule is placed on the switch flow table. However, the Proxy still receives the packets from that client.
How can I debug and fix this ?
Below my is  topology! 



              Controller Floodlight
          _________|________
          |                 |               |
    Netfpga ____OVS ___Netfpga
          |                 |               |
      SIP           Proxy         SIP
   Client1                         Client2


Thanks in advance, 
Best Regards,
Hssan



-- 
Documentation and tutorials: https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation
Posting guidelines: https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428
--- 
You received this message because you are subscribed to the Google Groups "Floodlight-developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to floodlight-dev+unsubscribe-XQarKSnW/Y+tkb4GaioC1Q@public.gmane.org.
To post to this group, send email to floodlight-dev-XQarKSnW/Y+tkb4GaioC1Q@public.gmane.org.
Visit this group at https://groups.google.com/a/openflowhub.org/group/floodlight-dev/.
To view this discussion on the web visit https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/CAJpa1GBbUxxHb8Ws7oEKiHx2drgU0AeEnN%3Dm0aRUFPjwb5OMnQ%40mail.gmail.com.

--
Documentation and tutorials: https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation
Posting guidelines: https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428
---
You received this message because you are subscribed to the Google Groups "Floodlight-developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to floodlight-dev+unsubscribe-XQarKSnW/Y+tkb4GaioC1Q@public.gmane.org.
To post to this group, send email to floodlight-dev-XQarKSnW/Y+tkb4GaioC1Q@public.gmane.org.
Visit this group at https://groups.google.com/a/openflowhub.org/group/floodlight-dev/.
To view this discussion on the web visit https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/DEEDC958-C00F-4D1E-9993-EF34FB71D057%40g.clemson.edu.

<div>
<div dir="ltr">
<div>
<div>
<div>hey Ryan , <br><br>
</div>You means when using dpctl add-flow ? <br>
</div>Do you know why ACL command doesn't get to block the port of that ip address ? <br><br>
</div>Thanks in advance&nbsp; <br><div><div><div>
<br><br>
</div></div></div>
</div>
<div class="gmail_extra">
<br><div class="gmail_quote">On Mon, Jun 27, 2016 at 5:05 PM, Ryan Izard <span dir="ltr">&lt;<a href="mailto:rizard <at> g.clemson.edu" target="_blank">rizard <at> g.clemson.edu</a>&gt;</span> wrote:<br><blockquote class="gmail_quote">
<div>You need to specify the dl_type (ethertype) as IPv4 (unless you're using IPv6, in which case it'll be IPv6).<div>
<br><div>
<div><div><span>Ryan</span></div></div>

</div>
<br><div><blockquote type="cite">
<div><div class="h5">
<div>On Jun 27, 2016, at 9:54 AM, Hssan Driss &lt;<a href="mailto:hssan.driss@..." target="_blank">hssan.driss@...</a>&gt; wrote:</div>
<br>
</div></div>
<div>
<div><div class="h5">
<div dir="ltr">
<div>Hi everyone,<span>&nbsp;</span><br>
</div>
<div>
<br>Yes, as a matter of fact, on Floodlight ACL choosing UDP protocol is translated into 11. Is that decimal or hex ?<span>&nbsp;</span><br>There might be a confusion though ! 0x11 = 17 (dec)<span>&nbsp;</span><br>
</div>And Since I'm using SIP on default 5060 port shall I precise it ?<div>
<div><br></div>
<div>In addition, I tried adding the flow on the switch manually using dpctl commands as below :<span>&nbsp;</span><br><br><span># sudo dpctl add-flow unix:/var/run/dp0.sock in_port=*,dl_vlan=*,dl_src=*,dl_dst=*,dl_type=*,nw_src=192.168.2.2,nw_dst=*,nw_proto=17,tp_src=*,tp_dst=5060,icmp_type=*,icmp_code=*,actions=<br><br></span>
</div>
<div><span><span>I get :<span>&nbsp;</span></span><br><br># sudo dpctl dump-flows unix:/var/run/dp0.sock<br>stats_reply (xid=0x54e91f95): flags=none type=1(flow)<br>&nbsp; cookie=0, duration_sec=1s, duration_nsec=552000000s, table_id=0, priority=32768, n_packets=0, n_bytes=0, idle_timeout=60,hard_timeout=0,actions=<br><br></span></div>
<div><span>And packets are blocked ! but can't get why the flow rule doesn't exactly match the rule added&nbsp;!<br><br></span></div>
<div><span>Thanks in advance for your help !<span>&nbsp;</span><br><br></span></div>
<div><span>Kind Regards,<span>&nbsp;</span><br></span></div>
<div><span>Hssan<span>&nbsp;</span><br></span></div>
<div><br></div>
</div>
</div>
<div class="gmail_extra">
<br><div class="gmail_quote">On Fri, Jun 24, 2016 at 4:25 PM, Hssan Driss<span>&nbsp;</span><span dir="ltr">&lt;<a href="mailto:hssan.driss@..." target="_blank">hssan.driss <at> gmail.com</a>&gt;</span><span>&nbsp;</span>wrote:<br><blockquote class="gmail_quote"><div dir="ltr"><div>
<div>
<div>Hi everyone,<br><br>I'm working on SDN lab. When intend to block SIP Client IP. We are pushing the blocking order based on the Src-ip address to Floodlight controller through REST API. Which is translated on a switch level :<br><br># dpctl dump-flows unix:/var/run/dp0.sock<br>stats_reply (xid=0x79853a73): flags=none type=1(flow)<br><span><span>&nbsp;<span>&nbsp;</span>cookie=45035999707555269, duration_sec=9s, duration_nsec=451000000s, table_id=0, priority=29999, n_packets=0, n_bytes=0, idle_timeout=0,hard_timeout=0,ip,nw_src=192.168.2.2,nw_proto=11,tp_src=0,tp_dst=0,actions=</span></span><br>&nbsp;<span>&nbsp;</span>cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=514, n_bytes=231936, idle_timeout=5,hard_timeout=0,in_port=4,dl_src=00:21:9b:88:04:66,dl_dst=18:03:73:b5:78:4d,actions=output:3<br>&nbsp;<span>&nbsp;</span>cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=685, n_bytes=383282, idle_timeout=5,hard_timeout=0,in_port=3,dl_src=18:03:73:b5:78:4d,dl_dst=00:21:9b:88:04:66,actions=output:4<br><br>
</div>
<div>SIP Client is on 192.168.2.2<br>
</div>
<br>The flow rule is placed on the switch flow table. However, the Proxy still receives the packets from that client.<br>
</div>How can I debug and fix this ?<br>Below my is&nbsp; topology!<span>&nbsp;</span><br><div>
<div>
<br><br><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>Controller Floodlight<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>_________|________<br><div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>|&nbsp;&nbsp;&nbsp;&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br>&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>Netfpga ____OVS ___Netfpga<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br>
</div>
<div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span>&nbsp;</span>SIP&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Proxy&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp; SIP<br>
</div>
<div>&nbsp;&nbsp;<span>&nbsp;</span>Client1 &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Client2<br>
</div>
<br>
</div>
<div>
<br>Thanks in advance,<span>&nbsp;</span><br>
</div>
<div>Best Regards,<br>
</div>
<div>Hssan<br>
</div>
<div><br></div>
</div>
</div></div></blockquote>
</div>
<br>
</div>
<div><br></div>
</div></div>
<span class=""><span>--<span>&nbsp;</span></span><br><span>Documentation and tutorials:<span>&nbsp;</span></span><a href="https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation" target="_blank">https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation</a><br><span>Posting guidelines:<span>&nbsp;</span></span><a href="https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428" target="_blank">https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428</a><br><span>---<span>&nbsp;</span></span><br><span>You received this message because you are subscribed to the Google Groups "Floodlight-developers" group.</span><br><span>To unsubscribe from this group and stop receiving emails from it, send an email to<span>&nbsp;</span></span><a href="mailto:floodlight-dev+unsubscribe@..." target="_blank">floodlight-dev+unsubscribe@...</a><span>.</span><br><span>To post to this group, send email to<span>&nbsp;</span></span><a href="mailto:floodlight-dev@..." target="_blank">floodlight-dev@...</a><span>.</span><br><span>Visit this group at<span>&nbsp;</span></span><a href="https://groups.google.com/a/openflowhub.org/group/floodlight-dev/" target="_blank">https://groups.google.com/a/openflowhub.org/group/floodlight-dev/</a><span>.</span><br></span><span>To view this discussion on the web visit<span>&nbsp;</span></span><a href="https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/CAJpa1GBbUxxHb8Ws7oEKiHx2drgU0AeEnN%3Dm0aRUFPjwb5OMnQ%40mail.gmail.com?utm_medium=email&amp;utm_source=footer" target="_blank">https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/CAJpa1GBbUxxHb8Ws7oEKiHx2drgU0AeEnN%3Dm0aRUFPjwb5OMnQ%40mail.gmail.com</a><span>.</span>
</div>
</blockquote></div>
<br>
</div>
</div>
<span class="">

<p></p>

-- <br>
Documentation and tutorials: <a href="https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation" target="_blank">https://floodlight.atlassian.net/wiki/display/floodlightcontroller/Floodlight+Documentation</a><br>
Posting guidelines: <a href="https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428" target="_blank">https://floodlight.atlassian.net/wiki/pages/viewpage.action?pageId=24805428</a><br>
--- <br>
You received this message because you are subscribed to the Google Groups "Floodlight-developers" group.<br>
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:floodlight-dev+unsubscribe@..." target="_blank">floodlight-dev+unsubscribe@...</a>.<br>
To post to this group, send email to <a href="mailto:floodlight-dev <at> openflowhub.org" target="_blank">floodlight-dev@...</a>.<br>
Visit this group at <a href="https://groups.google.com/a/openflowhub.org/group/floodlight-dev/" target="_blank">https://groups.google.com/a/openflowhub.org/group/floodlight-dev/</a>.<br></span>
To view this discussion on the web visit <a href="https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/DEEDC958-C00F-4D1E-9993-EF34FB71D057%40g.clemson.edu?utm_medium=email&amp;utm_source=footer" target="_blank">https://groups.google.com/a/openflowhub.org/d/msgid/floodlight-dev/DEEDC958-C00F-4D1E-9993-EF34FB71D057%40g.clemson.edu</a>.<br>
</blockquote>
</div>
<br>
</div>
</div>
Hssan Driss | 24 Jun 17:25 2016
Picon
Gravatar

Flow Rule not applied on the Switch Level

Hi everyone,

I'm working on SDN lab. When intend to block SIP Client IP. We are pushing the blocking order based on the Src-ip address to Floodlight controller through REST API. Which is translated on a switch level :

# dpctl dump-flows unix:/var/run/dp0.sock
stats_reply (xid=0x79853a73): flags=none type=1(flow)
  cookie=45035999707555269, duration_sec=9s, duration_nsec=451000000s, table_id=0, priority=29999, n_packets=0, n_bytes=0, idle_timeout=0,hard_timeout=0,ip,nw_src=192.168.2.2,nw_proto=11,tp_src=0,tp_dst=0,actions=
  cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=514, n_bytes=231936, idle_timeout=5,hard_timeout=0,in_port=4,dl_src=00:21:9b:88:04:66,dl_dst=18:03:73:b5:78:4d,actions=output:3
  cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=685, n_bytes=383282, idle_timeout=5,hard_timeout=0,in_port=3,dl_src=18:03:73:b5:78:4d,dl_dst=00:21:9b:88:04:66,actions=output:4

SIP Client is on 192.168.2.2

The flow rule is placed on the switch flow table. However, the Proxy still receives the packets from that client.
How can I debug and fix this ?
Below my is  topology!



              Controller Floodlight
          _________|________
          |                 |               |
    Netfpga ____OVS ___Netfpga
          |                 |               |
      SIP           Proxy         SIP
   Client1                         Client2


Thanks in advance,
Best Regards,
Hssan

<div><div dir="ltr"><div>
<div>
<div>Hi everyone,<br><br>I'm working on SDN lab. When intend to block SIP Client IP. We are pushing the blocking order based on the Src-ip address to Floodlight controller through REST API. Which is translated on a switch level :<br><br># dpctl dump-flows unix:/var/run/dp0.sock<br>stats_reply (xid=0x79853a73): flags=none type=1(flow)<br><span><span>&nbsp; cookie=45035999707555269, duration_sec=9s, duration_nsec=451000000s, table_id=0, priority=29999, n_packets=0, n_bytes=0, idle_timeout=0,hard_timeout=0,ip,nw_src=192.168.2.2,nw_proto=11,tp_src=0,tp_dst=0,actions=</span></span><br>&nbsp; cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=514, n_bytes=231936, idle_timeout=5,hard_timeout=0,in_port=4,dl_src=00:21:9b:88:04:66,dl_dst=18:03:73:b5:78:4d,actions=output:3<br>&nbsp; cookie=4503599627370496, duration_sec=39s, duration_nsec=677000000s, table_id=0, priority=100, n_packets=685, n_bytes=383282, idle_timeout=5,hard_timeout=0,in_port=3,dl_src=18:03:73:b5:78:4d,dl_dst=00:21:9b:88:04:66,actions=output:4<br><br>
</div>
<div>SIP Client is on 192.168.2.2<br>
</div>
<br>The flow rule is placed on the switch flow table. However, the Proxy still receives the packets from that client.<br>
</div>How can I debug and fix this ?<br>Below my is&nbsp; topology! <br><div>
<div>
<br><br><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Controller Floodlight<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; _________|________<br><div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br>&nbsp;&nbsp;&nbsp; Netfpga ____OVS ___Netfpga<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br>
</div>
<div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; SIP&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Proxy&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp; SIP<br>
</div>
<div>&nbsp;&nbsp; Client1 &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Client2<br>
</div>
<br>
</div>
<div>
<br>Thanks in advance, <br>
</div>
<div>Best Regards,<br>
</div>
<div>Hssan<br>
</div>
<div><br></div>
</div>
</div></div></div>
Christopher Hannon | 15 Jun 16:41 2016

Fast Failover question

[Question] From my understanding the Fast fail over option in group tables allows for monitoring a link and when the detected condition occurs will apply that group to the packets. This seems fast and would work well for a topology like

 h1 --- x --- x ---- x --- h2
           \           /
            x  ---  x

if the green path is the original

and a failure

 h1 --- x --- x ---- x --- h2
           \           /
            x  ---  x

equipped with FF could have a backup path:

 h1 --- x --- x ---- x --- h2
                    /
            x  ---  x

However if instead the failure is here:
 
 h1 --- x --- x ---- x --- h2
           \           /
            x  ---  x

then FF really wont help and we need to use a controller application to dynamically re-route traffic.

My question is: 

Is it possible for a switch to take an action when a link goes down such as inform neighboring switches or take a port down automatically or is this something that should be solved differently?

thanks,
<div><div dir="ltr">[Question] From my understanding the Fast fail over option in group 
tables allows for monitoring a link and when the detected condition 
occurs will apply that group to the packets. This seems fast and would 
work well for a topology like<br><br><span>&nbsp;h1 --- x --- x ---- x --- h2</span><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; \&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; /<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; x&nbsp; ---&nbsp; x<br><br><div>if the green path is the original<br><br>
</div>
<div>and a failure <br><br><span>&nbsp;h1 --- x<span> --- </span>x ---- x --- h2</span><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; \&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; /<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; x&nbsp; ---&nbsp; x<br><br>
</div>
<div>equipped with FF could have a backup path:<br><br><span>&nbsp;h1 --- x <span>--- x ---- </span>x --- h2</span><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span> \&nbsp;</span>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<span> /</span><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span>x&nbsp; ---&nbsp; x<br><span><br></span></span>
</div>
<div>
<span><span>However if instead the failure is here: </span><br></span>&nbsp;<br><span>&nbsp;h1 --- x --- x <span>----</span> x --- h2</span><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; \&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; /<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; x&nbsp; ---&nbsp; x<br><br>
</div>
<div>then FF really wont help and we need to use a controller application to dynamically re-route traffic.<br><br>
</div>
<div>My question is:&nbsp; <br><br>Is
 it possible for a switch to take an action when a link goes down such 
as inform neighboring switches or take a port down automatically or is 
this something that should be solved differently?<br><br>
</div>thanks, </div></div>
PUZZLE | 30 May 18:42 2016
Picon

Meaning of: JAVA_HOME not set, results may vary

Hi,
while using ODL Beryllium-SR2, I was seeing this message:

karaf: JAVA_HOME not set; results may vary
Java HotSpot(TM) 64-Bit Server VM warning: ignoring option MaxPermSize=512m; support was removed in 8.0

I did everything to define JAVA_HOME, and it was well defined in my ubuntu, but the message kept appearing, and After asking i was told to ignore it cause it hasn't really an effect,
Now after processing and plotting the throughput test results, i got this:


​which sounds to me very ILLOGICAL, is it because of that "results may vary" message??
Please help me to clarify

Tahnks
<div><div dir="ltr">
<div>Hi,</div>
<div>while using ODL&nbsp;Beryllium-SR2, I was seeing this message:</div>
<div><br></div>
<div>karaf: JAVA_HOME not set; results may vary</div>
<div>Java HotSpot(TM) 64-Bit Server VM warning: ignoring option MaxPermSize=512m; support was removed in 8.0</div>
<div><br></div>
<div>I did everything to define JAVA_HOME, and it was well defined in my ubuntu, but the message kept appearing, and After asking i was told to ignore it cause it hasn't really an effect,</div>
<div>Now after processing and plotting the throughput test results, i got this:</div>
<div><br></div>
<div>
<br>&#8203;which sounds to me very ILLOGICAL, is it because of that "results may vary" message??<br>
</div>
<div>Please help me to clarify</div>
<div><br></div>
<div>Tahnks</div>
</div></div>
PUZZLE | 26 May 03:19 2016
Picon

[PLOT] Cbench output

Hi,
I am trying to plot the cbench results output but i still didn't quite understood it especially due to the lack of documenttion,

Would anyone please help me understand the numbers so i can can decide what to plot?

Thank you very much!
<div><div dir="ltr">Hi,<div>I am trying to plot the cbench results output but i still didn't quite understood it especially due to the lack of documenttion,</div>
<div><br></div>
<div>Would anyone please help me understand the numbers so i can can decide what to plot?</div>
<div><br></div>
<div>Thank you very much!</div>
</div></div>
Haftay Gebreslasie Abreha | 11 May 10:59 2016
Picon

problem when using QoS with Flowvisor

Hello,

My name is Haftom. I am beginner for ubuntu as well as mininet. Currently, I want to do some research on SDN and NV. I have developed three nodes connected together which forms a triangle, s1, s2 and s3. And each switch get connected with two hosts i.e.  h1&h2 connected to s1 and h3&h4 connected to s2 and h5&h6 connected to s3. I have two opendaylight controllers (0.3.3-Lithium-SR3) installed on my PC with the ip address of 192.168.56.101 and .106. Moreover, I have Flowvisor v.1.4 installed on my PC which has an IP address of 192.168.56.103.

My aim is to have a research slice which containes only h1,h3 &h5 and will use only 1Mbps and get controlled by the controller with the ip address of 192.168.56.101. And I want to have a production slice which contains only h2,h4 and h6 and will use only 3Mbps and get controlled by the controller with the ip address of 192.168.56.106.  Then let me explain what I did.

1. I created two slices named ' research ' and ' production '

fvctl -f /dev/null add-slice research tcp:192.168.56.101:6633 admin <at> researchslice
fvctl -f /dev/null add-slice production tcp:192.168.56.106:6653 admin <at> productionslice

2. I created two queues q0 and q1 for each port like for example the first one is

ovs-vsctl set port s1-eth1 qos= <at> newqos -- --id= <at> newqos create qos type=linux-htb queues=0= <at> q0,1= <at> q1 -- --id= <at> q0 create queue other-config:min-rate=3000000 other-config:max-rate=3000000 -- --id= <at> q1 create queue other-config:min-rate=1000000 other-config:max-rate=1000000

3. I created the flowspace as follows

for the research slice 
fvctl -f /dev/null add-flowspace dpid1-port1 1 1 in_port=1 research=7   [h1 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid1-port3 1 1 in_port=3 research=7 
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid1-port4 1 1 in_port=4 research=7 
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid2-port1 2 1 in_port=1 research=7 
fvctl -f /dev/null add-flowspace dpid2-port2 2 1 in_port=2 research=7   [h3 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid2-port4 2 1 in_port=4 research=7 
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid3-port1 3 1 in_port=1 research=7 
fvctl -f /dev/null add-flowspace dpid3-port2 3 1 in_port=2 research=7   [h5 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=1 dpid3-port4 3 1 in_port=4 research=7      
 
 for the production slice
fvctl -f /dev/null add-flowspace dpid11-port2 1 1 in_port=2 production=7   [h2 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid11-port3 1 1 in_port=3 production=7 
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid11-port4 1 1 in_port=4 production=7 
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid22-port1 2 1 in_port=1 production=7
fvctl -f /dev/null add-flowspace dpid22-port3 2 1 in_port=3 production=7   [h4 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid22-port4 2 1 in_port=4 production=7 
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid33-port1 3 1 in_port=1 production=7 
fvctl -f /dev/null add-flowspace   dpid33-port3 3 1 in_port=3 production=7   [h6 is connected to this port]
fvctl -f /dev/null add-flowspace --forced-queue=0 dpid33-port4 3 1 in_port=4 production=7
     
4.  one of the controllers, flow-visor and mini-net are running
then I run the following command on mininet

 sudo mn --custom mininet/custom/Research.py --topo Research --mac --controller=remote,ip=192.168.56.103 --switch ovs,protocols=OpenFlow10 --link tc

5. However, it always uses the q0 rather than q1 although It is expected to use q1. After that I added the following for each port which are connected to the switches for example.

sudo ovs-ofctl add-flow s1 in_port=3, dl_type=0x0806, nw_proto=6, nw_dst=10.0.0.1, priority=100, actions=set_queue:1, normal

However, the flowvisor stops immediately.  But I can see the traffic is going thruogh q1 when I am not using the flowvisor meaning when I use this commands

 sudo mn --custom mininet/custom/Research.py --topo Research --mac --controller=remote,ip=192.168.56.101 --switch ovs,protocols=OpenFlow10 --link tc


can you please help me how to solve this and run my two controllers simultaneously without interfering the two slices by using q0 and q1?

Thank you


Regards,
Haftom
<div><div dir="ltr">Hello,<br><br>My name is Haftom. I am beginner for ubuntu as well as mininet. Currently, I want to do some research on SDN and NV. I have developed three nodes connected together which forms a triangle, s1, s2 and s3. And each switch get connected with two hosts i.e.&nbsp; h1&amp;h2 connected to s1 and h3&amp;h4 connected to s2 and h5&amp;h6 connected to s3. I have two opendaylight controllers (0.3.3-Lithium-SR3) installed on my PC with the ip address of 192.168.56.101 and .106. Moreover, I have Flowvisor v.1.4 installed on my PC which has an IP address of 192.168.56.103. <br><br>My aim is to have a research slice which containes only h1,h3 &amp;h5 and will use only 1Mbps and get controlled by the controller with the ip address of 192.168.56.101. And I want to have a production slice which contains only h2,h4 and h6 and will use only 3Mbps and get controlled by the controller with the ip address of 192.168.56.106.&nbsp; Then let me explain what I did.<br><br>1. I created two slices named ' research ' and ' production '<br><br>fvctl -f /dev/null add-slice research tcp:<a href="http://192.168.56.101:6633">192.168.56.101:6633</a> admin <at> researchslice<br>fvctl -f /dev/null add-slice production tcp:<a href="http://192.168.56.106:6653">192.168.56.106:6653</a> admin <at> productionslice<br><br>2. I created two queues q0 and q1 for each port like for example the first one is<br><br>ovs-vsctl set port s1-eth1 qos= <at> newqos -- --id= <at> newqos create qos type=linux-htb queues=0= <at> q0,1= <at> q1 -- --id= <at> q0 create queue other-config:min-rate=3000000 other-config:max-rate=3000000 -- --id= <at> q1 create queue other-config:min-rate=1000000 other-config:max-rate=1000000<br><br>3. I created the flowspace as follows<br><br>for the research slice&nbsp;
<br>fvctl -f /dev/null add-flowspace dpid1-port1 1 1 in_port=1 research=7&nbsp;&nbsp; [h1 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid1-port3 1 1 in_port=3 research=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid1-port4 1 1 in_port=4 research=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid2-port1 2 1 in_port=1 research=7&nbsp;
<br>fvctl -f /dev/null add-flowspace dpid2-port2 2 1 in_port=2 research=7&nbsp;&nbsp; [h3 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid2-port4 2 1 in_port=4 research=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid3-port1 3 1 in_port=1 research=7&nbsp;
<br>fvctl -f /dev/null add-flowspace dpid3-port2 3 1 in_port=2 research=7&nbsp;&nbsp; [h5 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=1 dpid3-port4 3 1 in_port=4 research=7
&nbsp;&nbsp;&nbsp; &nbsp;<br>&nbsp;<br>&nbsp;for the production slice<br>fvctl -f /dev/null add-flowspace dpid11-port2 1 1 in_port=2 production=7&nbsp;&nbsp; [h2 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid11-port3 1 1 in_port=3 production=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid11-port4 1 1 in_port=4 production=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid22-port1 2 1 in_port=1 production=7
<br>fvctl -f /dev/null add-flowspace dpid22-port3 2 1 in_port=3 production=7&nbsp;&nbsp; [h4 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid22-port4 2 1 in_port=4 production=7&nbsp;
<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid33-port1 3 1 in_port=1 production=7&nbsp;
<br>fvctl -f /dev/null add-flowspace&nbsp;&nbsp; dpid33-port3 3 1 in_port=3 production=7&nbsp;&nbsp; [h6 is connected to this port]<br>fvctl -f /dev/null add-flowspace --forced-queue=0 dpid33-port4 3 1 in_port=4 production=7
<br>&nbsp;&nbsp;&nbsp; &nbsp;<br>4.&nbsp; one of the controllers, flow-visor and mini-net are running<br>then I run the following command on mininet<br><br>&nbsp;sudo mn --custom mininet/custom/Research.py --topo Research --mac --controller=remote,ip=192.168.56.103 --switch ovs,protocols=OpenFlow10 --link tc
<br><br>5. However, it always uses the q0 rather than q1 although It is expected to use q1. After that I added the following for each port which are connected to the switches for example.<br><br>sudo ovs-ofctl add-flow s1 in_port=3, dl_type=0x0806, nw_proto=6, nw_dst=10.0.0.1, priority=100, actions=set_queue:1, normal<br><br>However, the flowvisor stops immediately.&nbsp; But I can see the traffic is going thruogh q1 when I am not using the flowvisor meaning when I use this commands<br><br>&nbsp;sudo mn --custom mininet/custom/Research.py --topo Research --mac --controller=remote,ip=192.168.56.101 --switch ovs,protocols=OpenFlow10 --link tc
<br><br><br>can you please help me how to solve this and run my two controllers simultaneously without interfering the two slices by using q0 and q1?<br><br>Thank you<br><br><br>Regards,<br>Haftom<br>
</div></div>
Devin Roberson | 10 May 16:31 2016

Translate between IPv4 and IPv6 packets

Hello,

I'm trying to forward packets between an IPv4 and an IPv6 network using 
an
OpenFlow switch. For this, I have to replace the IPv4 header with an 
IPv6 header
(and the other way around). I have tried using Open vSwitch and doing a
set-field action on the dl_type field 
(actions=set_field:0x86dd->dl_type), but I
get an error that the field is read-only. Is this a limitation in Open 
vSwitch
or in the OpenFlow specification? Or is there perhaps another way of
accomplishing this? (without forwarding each packet to a controller and
rewriting them there.)

Thanks in advance.
Slavica Tomovic | 9 May 12:11 2016
Picon

OpenvSwitch

Hello experts!

I want to implement OpenFlow wireless networks by using Raspberry Pi
devices. My idea is to install OVS software on them. Is it possible to
such a thing?

I read that OVS cannot be used for wireless networks, but what if I
add multiple wi-fi interfaces to each RPI and interconnect them
wirelessly?

Also, is it maybe possible to use only one physical wifi NIC and then
configure multiple virtual interfaces on it?

I just need information (ASAP) is it possible (and how) to use OVS in
the data-plane of OpenFlow wireless network.

Thanks in advance,

Slavica
fatt 3issam | 3 May 22:12 2016
Picon

OFPT_QUEUE_GET_CONFIG_REPLY issue

I want to get the properties of a queue set on an OVS port, but all i get via the OFPT_QUEUE_GET_CONFIG_REPLY is: the queue id, the port number, length and pad!!

The queue is correctly set and here is it's config:
# ovs-vsctl set port switch-port qos= <at> newqos -- --id= <at> newqos create qos type=linux-htb other-config:max-rate=20000000 queues:0= <at> newqueue -- --id= <at> newqueue create queue other-config:min-rate=1000000 other-config:max-rate=1000000

So is there anything to do to get those rates?
<div><div dir="ltr">I want to get the properties of a queue set on an OVS port, but all i get via the OFPT_QUEUE_GET_CONFIG_REPLY is: the queue id, the port number, length and pad!!<div><br></div>
<div>The queue is correctly set and here is it's config:</div>
<div><span># ovs-vsctl set port switch-port qos= <at> newqos -- --id= <at> newqos create qos type=linux-htb other-config:max-rate=20000000 queues:0= <at> newqueue -- --id= <at> newqueue create queue other-config:min-rate=1000000 other-config:max-rate=1000000</span></div>
<div><span><br></span></div>
<div>So is there anything to do to get those rates?</div> 		 	   		  </div></div>
Hssan Driss | 28 Apr 17:23 2016
Picon
Gravatar

connecting host to the controller northbound

Hi everyone,

I' m working on an SDN plateform using a controller machine and 3 Switches as below. I'm using floodlight as a switch.

h1   h2   h3
|       |       |
s1---s2---s3
|       |       |
|------c-----|

I need to use a rest client from a host (h1) as an application to communicate with the controller's REST API.

Any suggestions on how to do this ?


Best Regards,
Hssan
<div><div dir="ltr">
<div>
<div>Hi everyone, <br><br>
</div>I' m working on an SDN plateform using a controller machine and 3 Switches as below. I'm using floodlight as a switch. <br><br>
</div>h1 &nbsp; h2 &nbsp; h3<br><div>|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&nbsp;&nbsp;&nbsp; &nbsp;&nbsp; |<br><span class=""><span class="">s1---s2---s3<br></span></span>
</div>
<div><span class=""><span class="">|&nbsp; &nbsp; &nbsp;&nbsp; |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br></span></span></div>
<div><span class=""><span class="">|------c-----|<br><br></span></span></div>
<div><span class=""><span class="">I need to use a rest client from a host (h1) as an application to communicate with the controller's REST API. <br><br>Any suggestions on how to do this ?<br><br><br></span></span></div>
<div><span class=""><span class="">Best Regards, <br></span></span></div>
<div>
<span class=""><span class="">Hssan</span></span> <br>
</div>
</div></div>

Gmane