On 05/19/2017 03:22 AM, Faber wrote:
i think there is an issue with my patch; i'm following the
original idea
to have stage ip similar to prod ip
for example:
noc01.stg.phx2.fedoraproject.org -> 10.5.126.2
noc01-a.stg.phx2.fedoraproject.org -> 10.5.128.41 [ matches noc01 on
10.5.126.41]
Note that we want them to have the same exact names as they do now. This
is because we have our staging env set in /etc/resolv.conf to search for
hosts in
stg.phx2.fedoraproject.org first... so in staging a 'noc01'
would get the ip for 'noc01.stg.phx2.fedoraproject.org' but in
production would get 'noc01.phx2.fedoraproject.org' We do this so we can
keep a lot of the config the same and it just works.
btw with dhcp IPs going from 10.5.128.10 to 10.5.128.20 there will be
duplicates
@smooge is it ok to use your
stg.phx2.fedoraproject.org file but with
the 128 ip space instead of 126?
so we have ips moved from 10.5.126 -> 10.5.128 numbers, but not 1:1
While it might be handy to make them 1:1 on IP's, I don't think it's a
requirement... so sure.
it seems all IPs there are unique, except for fas-all and fas01
both
with 10.5.126.86
Yep. That fas-all in stg is just a pointer to the one fas server (since
we only have one there). In production it points to all 3 of the servers
we have in place.
kevin