> On Feb. 18, 2013, 7:59 a.m., Koushik Das wrote:
> > plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/VmwareServerDiscoverer.java,
> >  line 185
> > <https://reviews.apache.org/r/9210/diff/1/?file=254475#file254475line185>
> >
> >     Is this behavior to default to svs discussed in the FS? If not please 
> > try to get a closure.

Makes sense to throw error in case of faulty vswitch type specified as 
parameter.
Now addCluster attempt is failed if errorneous vswitch type is supplied.


> On Feb. 18, 2013, 7:59 a.m., Koushik Das wrote:
> > plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/resource/VmwareResource.java,
> >  line 285
> > <https://reviews.apache.org/r/9210/diff/1/?file=254478#file254478line285>
> >
> >     Looks like _nexusVSwitch is no longer getting used

Yes, removed it.


- Sateesh


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/9210/#review16696
-----------------------------------------------------------


On Feb. 25, 2013, 10:06 a.m., Sateesh Chodapuneedi wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/9210/
> -----------------------------------------------------------
> 
> (Updated Feb. 25, 2013, 10:06 a.m.)
> 
> 
> Review request for cloudstack, Murali Reddy and Kelven Yang.
> 
> 
> Description
> -------
> 
> This is 5/final patch for feature 'Support for VMware dvSwitch in CloudStack'.
> 
> This patch contains 
> 1)Changes to addCluster done in vmware discoverer to support vswitch type 
> provided as parameters. Also performing validation of vswitch type parameter 
> provided with addCluster api call. Physical network configuration is 
> validated.
> 2)Changes to vmware resource to use specified vswitch type while preparing 
> network for guest and public traffic types
> 3)Changes to vmware manager to introduce new global parameter 
> vmware.ports.per.dvportgroup. Some cleanup.
> Virtual switch type could be chosen at zone level or at cluster level for 
> specific traffic type. autoExpand of dvPortGroup is available in code but 
> disabled as its breaking because vCenter 4.1 does not support autoExpand 
> feature. This would be supported once vSphere SDK 5.1 is added to CloudStack.
> 
> 
> This addresses bug CLOUDSTACK-657.
> 
> 
> Diffs
> -----
> 
>   
> plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/VmwareServerDiscoverer.java
>  5d7edce 
>   
> plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/manager/VmwareManager.java
>  445b2f0 
>   
> plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/manager/VmwareManagerImpl.java
>  70f98cc 
>   
> plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/resource/VmwareResource.java
>  5cac253 
>   server/src/com/cloud/resource/ResourceManagerImpl.java 98044fb 
>   vmware-base/src/com/cloud/hypervisor/vmware/mo/HypervisorHostHelper.java 
> 50f9541 
> 
> Diff: https://reviews.apache.org/r/9210/diff/
> 
> 
> Testing
> -------
> 
> Testing
> -------
> 
> Manual testing:-
> 1) Tested guest traffic over dvSwitch on a dedicated physical network. In 
> this case management and public traffic uses standard vSwitch on a common 
> physical network.
> 2) Tested both guest traffic and public traffic over dvSwitch on a physical 
> network.
> 3) Use optional parameters added to AddClusterCmd to override Zone level 
> network traffic label. Tested 2 clusters, one with standard vSwitch and other 
> with dvSwitch.
> 4) Tested all 3 traffic types on single physical network with global 
> parameter 'vmware.use.dvswitch' set to false. This is default configuration 
> scenario.
> 
> 
> Added following tests,
> 1) Test fetching dvSwitch object from vCenter
> 2) Test for presence of dvPortGroup
> 3) Test presence of dvPortGroup
> 4) Test get existing dvPortGroup
> 5) fetch dvPortGroup configuration
> 6) Test compare dvPortGroup configuration
> 7) Test update dvPortGroup configuration
> 
> 
> Thanks,
> 
> Sateesh Chodapuneedi
> 
>

Reply via email to