Re: [Kimchi-devel] [RFC] [Kimchi] Add support to edit virtual networks

On 20-04-2016 16:54, Socorro Stoppler wrote: >>>> >>>> *** FRONTEND *** >>>> >>>> * Add new option Edit to Actions menu under >>>> Virtualization/Network tab, enabled only for stopped networks >>>> * When clicked, this option will open a new window for >>>> editing >>>> that virtual network, with the following fields. Based on >>>> network type, some of below fields may be disabled: >>>> >>>> Type: just a label describing the network type, not >>>> changeable. >>>> Name: all types of virtual network can be renamed >>>> Interfaces: attach/detach interfaces (enabled for VEPA) or >>>> change interface (for macvtap, bridged) >>>> Subnet/DHCP range: two IP fields: start, end (enabled for >>>> NAT, >>>> isolated) >>>> VLAN id: integer field (enabled for bridged networks) >>>> >>>> >>>> I'm ok with this. The modal window should look almost the same as >>>> the "Add Network" window, except for the Network type that will >>>> be a >>>> "readonly" input. >>>> >>>> >>>> BTW, should we add a mask for the Subnet and DHCP IP fields? I >>>> think I >>>> saw somewhere someone asking for a mask for IP addresses but I wasn't >>>> able to find anything that supports both ipv4 and ipv6 strings at the >>>> same time. >>> >>> Hi Samuel, this is not correct, sorry for the confusion. >>> >>> It should be only one field. Call it "Address space", same term used >>> when listing the networks. >>> >>> That field will accept something like 127.0.0.0/24 > Which of the IP address do you want to be displayed? Doesn't the > backend provide both subnet and dhcp in some cases? In my opinion only subnet should be displayed. The dhcp will be automatically changed by backend if subnet is changed by the user. We could display dhcp range (start,end) as well, but if the user changes subnet, there will be no reflection on the dhcp range values.

>>> >>> I'm not sure about IPV6 support in Kimchi. Aline? >>> >> >> It would be good to have a mask for the fields to restrict user >> inputs. Only IPv4 is supported in the current implementation.