Changes between Version 2 and Version 3 of HowTo/ShareALanOnAHWSwitch
- Timestamp:
- 07/29/14 10:03:55 (10 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
HowTo/ShareALanOnAHWSwitch
v2 v3 6 6 7 7 1) Reserve a VM and start your controller. I don't think it matters much which aggregate you pick for MYAGG. In fact, you don't need a GENI node at all as you could run the controller on your own machine. 8 {{{ 8 9 omni createsliver MYSLICE http://www.gpolab.bbn.com/exp/OpenFlowOVS/pox-controller.rspec -a MYAGG --useSliceMembers 10 }}} 9 11 2) Edit the attached max-stanford-stitched-flack.rspec to point to the controller. The controller should be accessed via the public IP of the node from step (1). 10 12 3) Reserve a VM at Stanford connected to a VM at MAX via a stitched link. These nodes will be "host1" & "host2": 13 {{{ 11 14 stitcher createsliver MYSLICE2 max-stanford-stitched-flack.rspec --useSliceMembers 15 }}} 12 16 4) Use performOperationalAction (poa) to share the stitched VLAN at Stanford. (Detailed instructions are in [2].) 13 17 a) Make a file called lan_options.json containing the following text: … … 20 24 (NOTE: "stitched0" should be the name of the stitched VLAN from the previous step and "ofstitchedlan" should be the name you give the shared VLAN and is used by the RSpec in step (4)): 21 25 22 b) omni -V 3 -a stanford-ig poa MYSLICE2 geni_sharelan --optionsfile lan_options.json 23 4) Reserve a third VM on the shared VLAN: 24 omni createsliver MYSLICE3 host3-stanford.rspec -a stanford-ig --useSliceMembers 26 b) 27 {{{ 28 omni -V 3 -a stanford-ig poa MYSLICE2 geni_sharelan --optionsfile lan_options.json 29 }}} 30 4) Reserve a third VM (`host3`) on the shared VLAN: 31 {{{ 32 omni createsliver MYSLICE3 host3-on-shared-vlan.rspec -a stanford-ig --useSliceMembers 33 }}} 34 5) Reserve a fourth VM (`host4`) on the shared VLAN: 35 {{{ 36 omni createsliver MYSLICE3 host4-on-shared-vlan.rspec -a max-ig --useSliceMembers 37 }}} 25 38 26 39 You should have a controller running at an aggregate of your choice and controlling the hardware switch connecting the three VMs (two VMs at Stanford IG and one VM at MAX IG) which are all on the same shared VLAN.