[Non-DoD Source] Re: BIND Masters and slaves

[Kevin Darcy]

[ Classification Level: PUBLIC ]

My 2 cents...

The "master" nomenclature is appropriate from a *data*dependency*
standpoint. The "master" holds the "master copy" of the zone contents (
https://www.collinsdictionary.com/us/dictionary/english/master-copy). All
other copies are duplicates of that.

Where it is inaccurate, is from a *functional* perspective, describing
replication. The so-called "master" doesn't reach out and *force* the
"slaves" to replicate the zone. It can send a NOTIFY, but that's just a
suggestion. Whenever an entity can choose to act, or not act, then "slave"
is the wrong term to use. Real-life slaves had no such freedom.

DNS has always had this dual identity as both a database architecture and a
network connectivity architecture. So, I would say, keep the "master"
terminology only in the context of the database aspects of DNS, maybe
limited to combining phrases such as "master file" or "master copy". When
discussing replication, however, perhaps it is more appropriate to switch
to the primary/secondary nomenclature, or maybe even
replication-client/replication-server nomenclature.


                    - Kevin


On Mon, Jun 15, 2020 at 2:17 PM Michael De Roover <isc at nixmagic.com> wrote:

> Of course I could, but I do not feel like the effort to change
> nomenclature is either beneficial or worth taking for granted the requests
> of some people on Twitter - as the slave to peer authority I am - given how
> much it affects documentation, code, comments, general environment of the
> projects themselves. I enjoy being surrounded by people much smarter than I
> am when it comes to the mailing list here. Let's keep it that way and not
> derange ourselves into meaningless blabber from social media.
>
> What I did notice over time however that most of the projects affected are
> also those who do have to maintain a good public image, usually
> corporations. Meanwhile projects such as Opal
> <https://github.com/opal/opal/issues/941> and recently Rubocop
> <https://github.com/rubocop-hq/rubocop/issues/8091> as well were not. The
> latter one I'd like to draw attention to. The maintainer clearly didn't ask
> for this and asked everyone who shamed him, why are you doing this? None of
> the complainers were affiliated to the project at all. Chances are that
> they weren't even using it and just searched for projects with the name
> "cop" in it instead. These are not the people I want to support in my
> effort to end racism, which I *do* support, and quite heavily so.
> On 6/15/20 8:00 PM, DeCaro, James John (Jim) CIV DISA FE (USA) wrote:
>
> Or you can call the slave servers 'secondary' servers.
>
> --
> Met vriendelijke groet / Best regards,
> Michael De Roover
> _______________________________________________
> Please visit https://lists.isc.org/mailman/listinfo/bind-users to
> unsubscribe from this list
>
> ISC funds the development of this software with paid support
> subscriptions. Contact us at https://www.isc.org/contact/ for more
> information.
>
>
> bind-users mailing list
> bind-users at lists.isc.org
> https://lists.isc.org/mailman/listinfo/bind-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20200615/08e23375/attachment-0001.htm>