Fix the portal net used in the default ubuntu cluster scripts...

[a-robinson]

...to actually be on a private IP range.

This came up in issue #6667.

[roberthbailey]

LGTM.

[thockin]

LGTM but I did not audit the rest of the ubuntu scripts to ensure that is unique... did you?

[a-robinson]

Ugh, you're completely right, I don't actually know how flannel is allocating pod IPs in these setups, so I don't know whether they'll conflict. I should definitely learn how, but don't have the time today.

In the meantime, this shouldn't be merged.

[WIZARD-CXY]

@a-robinson The ip range that flannel used is configured by using `/opt/bin/etcdctl mk /coreos.com/network/config '{"Network":"10.0.0.0/16"}' by user
And our team is working on merge the ubuntu single node and multiple nodes into one as this pr
#5498.

[a-robinson]

Ah, I wasn't aware of that work, @WIZARD-CXY. Thanks for the pointer! Is there any chance that, while you guys are in there, you or @resouer could swap out the 11.1.1.0/24 portal nets for a subnet in an actual private IP range, like something under 10.0.0.0/8 or 172.16.0.0/12?

[WIZARD-CXY]

@a-robinson ok no problem

[a-robinson]

Thanks!