Version 35 (modified by 12 years ago) (diff) | ,
---|
EG-EXP-5: ExoGENI OpenFlow Network Resources Acceptance Test
This page captures status for the test case EG-EXP-5, which verifies the ability to support OpenFlow operations and integration with meso-scale compute resources and other compute resources external to the ExoGENI rack. For overall status see the ExoGENI Acceptance Test Status page. Last Update: 08/07/12
Test Status
This section captures the status for each step in the acceptance test plan.
State Legend | Description |
Color(green,Pass)? | Test completed and met all criteria |
Color(#98FB98,Pass: most criteria)? | Test completed and met most criteria. Exceptions documented |
Color(red,Fail)? | Test completed and failed to meet criteria. |
Color(yellow,Complete)? | Test completed but will require re-execution due to expected changes |
Color(orange,Blocked)? | Blocked by ticketed issue(s). |
Color(#63B8FF,In Progress)? | Currently under test. |
Test Plan Steps
This test case is modified from the original test case in the ExoGENI Acceptance Test Plan, it combines the two experiments in to one experiment, so only one experimenter account is used. Although multiple slices are still used to run the experiment.
For this experiment these are OpenFlow network resource at the BBN site that are managed via the BBN ExoGENI FOAM:
- Two ExoGENI VMs on OF shared VLAN 1750
- Two BBN Campus VMs access OF VLAN 1750 via flowspaces through the ExoGENI 1OpenFlow switch.
Additionally non OpenFlow VLAN traffic is being exchanged via the ExoGENI OpenFlow Switch
- Two BBN campus VMs access non-OF Multi-point VLAN 3705 via flowspaces through the ExoGENI FOAM.
The remotes OpenFlow resources for VLAN 1750 include:
- Two meso-scale remote sites: 1 WAPG at Rutgers(Internet2) and 1 MyPLC at Clemson (NLR).
- NRL and Internet2 OpenFlow backbone VLAN 3716 is used for all hosts in experiment1 to exchange traffic
The remotes for the non-OF VLAN 3705 will include four remote end-points: Washington, Wisconsin, Indiana, and a BBN.
The pgeni.gpolab.bbn.com slice authority is used for the credentials and the following aggregate manager nick_names were defined in the omni_config:
exobbn=,https://bbn-hn.exogeni.net:11443/orca/xmlrpc of-exobbn=,https://bbn-hn.exogeni.net:3626/foam/gapi/1 of-bbn=,https://foam.gpolab.bbn.com:3626/foam/gapi/1 of-i2=,https://foam.net.internet2.edu:3626/foam/gapi/1 of-rutgers=,https://nox.orbit-lab.org:3626/foam/gapi/1 pg2=,https://www.emulab.net:12369/protogeni/xmlrpc/am/2.0 of-clemson=,https://foam.clemson.edu:3626/foam/gapi/1 plc-clemson=,http://myplc.clemson.edu:12346/
Step 1. As Experimenter1, Determine GPO compute resources and define RSpec.
Two local BBN Campus VMs (siovale and tiberium) have an interface for network 10.42.11.0/24 which will be used as the address range for the OF traffic. These nodes are NOT part of the ExoGENI rack, but will exchange traffic on VLAN 1750 via the rack OpenFlow Switch.
The two BBN Campus VMs (siovale and tiberium) also have a second interface on the 10.37.5.0/24 network range which is used for the non-OF VLAN 3705 traffic exchange.
Step 2. Determine remote meso-scale compute resources and define RSpec.
In experiment2 there are three mesoscale site, Rutgers, Clemson and BBN, the following hosts are selected in the RSpecs:
- 1 WAPG node (pg51) at Rutgers Site via Internet2
- 1 MyPLC node (planetlab4.clemson.edu) at Clemson Site via NLR
- 1 MyPLC node (navis.gpolab.bbn.com) at BBN Site Via NLR
Step 3. Define a request RSpec for OF network resources at the BBN ExoGENI FOAM.
The BBN ExoGENI OF traffic will access the OF backbone via the BBN OpenFlow switch poblano. An RSpec is required for the BBN OF switch to map the VLAN 1750 from the ExoGENI rack to the core VLAN 3716. The following FOAM RSPecs are need for the BBN site:
- EG-EXP-5-scenario1-openflow-exobbn.rspec - BBN ExoGENI Rack FOAM Aggregate network resource request RSpec.
- EG-EXP-5-scenario1-openflow-bbn.rspec - BBN FOAM Aggregate network resource request RSpec.
Step 4. Define a request RSpec for OF network resources at the remote I2 meso-scale site.
An RSpec was defined for each site FOAM aggregate and compute resources aggregate. The following RSpec were used for the meso-scale sites:
- EG-EXP-5-scenario1-wapg-rutgers.rspec - Rutgers WAPG nodes pg51 compute resource request RSpec.
- EG-EXP-5-scenario1-openflow-rutgers.rspec - Rutgers FOAM Aggregate network resource request RSpec.
- EG-EXP-5-scenario1-myplc-clemson.rspec - Clemson MyPLC node planetlab4 compute resource request RSpec.
- EG-EXP-5-scenario1-openflow-clemson.rspec - Clemson FOAM Aggregate network resource request RSpec.
- EG-EXP-5-scenario1-myplc-bbn.rspec - BBN MyPLC node navis compute resource request RSpec.
- EG-EXP-5-scenario1-openflow-bbn.rspec - BBN FOAM Aggregate network resource request RSpec.
Step 5. Define a request RSpec for the OpenFlow Core resources
The OF core resources for VLAN 3716 were defined as follows:
- EG-EXP-5-scenario1-openflow-i2.rspec = Internet2 Core FOAM Aggregate network resources request RSpec.
- EG-EXP-5-scenario1-openflow-nlr.rspec - NLR Core FOAM Aggregate network resources request RSpec.
Step 6. Create the first slice
A sliver was created for the following:
./src/omni.py createslice EG-EXP-5-scenario1
Step 7. Create a sliver for the GPO compute resources
The site resources were not reserved compute resource for the test case. These are standalone Linux VM hosts that exist on the vSwitch Karshyyyk, this test plan step was incorrect. This step is deemed complete by simply setting up the nodes with the right address for the VLAN 1750.
Step 8. Create a sliver at the I2 meso-scale site using FOAM at site
Created a sliver at the Rutgers FOAM for VLAN 3716:
./src/omni.py -a of-rutgers createsliver EG-EXP-5-scenario1 EG-EXP-5-scenario1-openflow-rutgers.rspec
Step 9. Create a sliver at of the BBN ExoGENI FOAM Aggregate
Created a sliver at the BBN ExoGENI FOAM:
./src/omni.py -a of-exobbn createsliver EG-EXP-5-scenario1 EG-EXP-5-scenario1-openflow-exobbn.rspec
Step 10. Create a sliver for the OpenFlow resources in the core.
Created a sliver at the I2 FOAM for VLAN 3716:
./src/omni.py -a of-i2 createsliver EG-EXP-5-scenario1 EG-EXP-5-scenario1-openflow-i2.rspec
Step 11. Create a sliver for the meso-scale compute resources.
The compute resource used for this test is pg51 a Wide Area PG node at Rutgers, and the sliver was created as follows:
./src/omni.py -a pg2 createsliver EG-EXP-5-scenario1 EG-EXP-5-scenario1-wapg-rutgers.rspec --api-version 2 -t GENI 3
Step 11. Log in to each of the compute resources and send traffic to the other end-point.
Logged into the WAPG node pg51:
$ ssh pg51.emulab.net Last login: Wed Jun 20 10:52:44 2012 from arendia.gpolab.bbn.com [lnevers@pg2 ~]$
Configure IP address to be used on VLAN 1750. First created the file /etc/sysconfig/network-scripts/ifcfg-eth1.1750 which contained the following:
DEVICE=eth1.1750 USERCTL=no VLAN=yes IPADDR=10.42.11.151 NETMASK=255.255.255.0 BOOTPROTO=static ONBOOT=yes
Started and checked the interface:
sudo ifup eth1.1750 ifconfig eth1.1750
Step 12. Verify that traffic is delivered to target.
Ping traffic delivered to each remote from each host in the experiment.
Step 13. Review baseline, GMOC, and meso-scale monitoring statistics.
Reviewed information all ports used in poblano RSpec that relate to the connections in this experiment:
Step 14. As Experimenter2, determine GPO compute resources and define RSpec.
This test case combined the two experiments into one experiment, thus this step is still executed by the first experimenter. Two local BBN Campus VMs (siovale and tiberium) are set up on the on vSwitch Karshyyyk. These nodes are NOT part of the ExoGENI rack.
Step 15. Determine remote meso-scale compute resources and define RSpec.
The remote meso-scale nodes to be used is planetlab4 a MyPLC node at Clemson.
Step 16. Define a request RSpec for OF network resources at the BBN ExoGENI FOAM.
Defined the following request RSpec for the Compute resource:
- EG-EXP-5-scenario1-myplc-clemson.rspec - MyPLC RSpec used to MyPLC compute resource.
Step 17. Define a request RSpec for OF network resources at the remote NLR meso-scale site.
Defined RSpec for the Clemson OpenFlow Sliver:
- EG-EXP-5-scenario1-openflow-clemson.rspec - Clemson FOAM Aggregate network resources request RSpec.
Step 18. Define a request RSpec for the OpenFlow Core resources
The OF core resources for VLAN 3716 were defined as follows:
- EG-EXP-5-scenario1-openflow-nlr.rspec - NLR Core FOAM Aggregate network resources request RSpec.
Step 19. Create the second slice
Note: In this test case multiple slices are used for the one experiment.
Created a second slice:
./src/omni.py createslice exobbnclemson
Step 20. Create a sliver for the GPO compute resources
Two local BBN Campus VMs (siovale and tiberium) are set up on the on vSwitch Karshyyyk. These nodes are NOT part of the ExoGENI rack and no reservations are need to used them. Test plan procedure is incorrect.
Step 21. Create a sliver at the meso-scale site using FOAM at site.
Because this run is being done with one experimenter, the second slice uses the Sliver from the first experiment, which set up the following:
./src/omni.py -a of-exobbn createsliver EG-EXP-5-scenario1 exorspec/EG-EXP-5-scenario1-openflow-exobbn.rspec
Step 22. Create a sliver at of the BBN ExoGENI FOAM Aggregate.
A sliver was created at the ExoGENI FOAM aggregate:
./src/omni.py -a of-bbn createsliver exobbnclemson EG-EXP-5-scenario-openflow-bbn.rspec
Step 23. Create a sliver for the OpenFlow resources in the core.
Reservation for NLR VLAN 3716:
./src/omni.py -a of-nlr createsliver exobbnclemson EG-EXP-5-scenario-openflow-nlr.rspec
Step 24. Create a sliver for the meso-scale compute resources.
Created a sliver for the FOAM aggregate and for the compute resources at Clemson:
./src/omni.py -a of-clemson createsliver exobbnclemson EG-EXP-5-scenario-openflow-clemson.rspec ./src/omni.py -a plc-clemson createsliver exobbnclemson EG-EXP-5-scenario-myplc-clemson.rspec
Step 25. Log in to each of the compute resources and send traffic to the other endpoint.
Log into the MyPLC host and ping the VLAN 1750 interfaces for siovale and tiberium:
$ ssh pgenigpolabbbncom_exobbnclemson@planetlab4.clemson.edu Last login: Wed Jun 20 17:42:41 2012 from arendia.gpolab.bbn.com [pgenigpolabbbncom_exobbnclemson@planetlab4 ~]$ ping 10.42.11.22 -c 5 PING 10.42.11.22 (10.42.11.22) 56(84) bytes of data. 64 bytes from 10.42.11.22: icmp_seq=1 ttl=64 time=164 ms 64 bytes from 10.42.11.22: icmp_seq=2 ttl=64 time=164 ms 64 bytes from 10.42.11.22: icmp_seq=3 ttl=64 time=165 ms 64 bytes from 10.42.11.22: icmp_seq=4 ttl=64 time=165 ms 64 bytes from 10.42.11.22: icmp_seq=5 ttl=64 time=165 ms --- 10.42.11.22 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4169ms rtt min/avg/max/mdev = 164.830/164.955/165.018/0.371 ms [pgenigpolabbbncom_exobbnclemson@planetlab4 ~]$ pgenigpolabbbncom_exobbnclemson@planetlab4 ~]$ ping 10.42.11.23 -c 5 PING 10.42.11.23 (10.42.11.23) 56(84) bytes of data. 64 bytes from 10.42.11.23: icmp_seq=1 ttl=64 time=164 ms 64 bytes from 10.42.11.23: icmp_seq=2 ttl=64 time=164 ms 64 bytes from 10.42.11.23: icmp_seq=3 ttl=64 time=165 ms 64 bytes from 10.42.11.23: icmp_seq=4 ttl=64 time=164 ms 64 bytes from 10.42.11.23: icmp_seq=5 ttl=64 time=165 ms --- 10.42.11.23 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4169ms rtt min/avg/max/mdev = 164.851/164.956/165.091/0.096 ms
Step 26. As Experimenter2, insert flowmods and send packet-outs only for traffic assigned to the slivers.
Will get a NOX controller after GEC, checked with Niki Riga who will provide the controller.
Step 27. Verify that traffic is delivered to target according to the flowmods settings.
Same as previous step.
Step 28. Review baseline, GMOC, and monitoring statistics.
Reviewed information for the ExoGENI port 15 on poblano for VLAN 1750 at:
Port 15 is the port used the connect the ExoGENI rack switch to BBN OF site switch, which has the connections to the backbone and to another OpenFlow switch at BBN (habanero) where the local meso-scale MyPLC host is connected.
Step 29. Stop traffic and delete slivers.
Stopped ping traffic and review the statistics. Issue a delete sliver at each aggregate, which has not been yet completed, trying to run experiment as long as possible before GEC.
Attachments (11)
- EG-EXP-5-exp1-measure.jpg (601.9 KB) - added by 12 years ago.
- EG-EXP-5-exp1-resources.jpg (242.1 KB) - added by 12 years ago.
- EG-EXP-5-exp1-slivers-EG-GPO.jpg (309.1 KB) - added by 12 years ago.
- EG-EXP-5-exp1.jpg (139.3 KB) - added by 12 years ago.
- EG-EXP-5-exp2.jpg (141.0 KB) - added by 12 years ago.
- EG-EXP-5-slices.jpg (117.4 KB) - added by 12 years ago.
- EG-EXP-5-exp1-part2.jpg (142.0 KB) - added by 12 years ago.
- EG-EXP-5-exp1-sliver-part2.jpg (154.3 KB) - added by 12 years ago.
- EG-EXP-5-slices-part2.jpg (129.2 KB) - added by 12 years ago.
- EG-EXP-5-sliver-urn-part2.jpg (276.6 KB) - added by 12 years ago.
- EG-EXP-sliver-measure.jpg (420.6 KB) - added by 12 years ago.