[Pdns-users] PowerDNS needs your thoughts on two important DNSSEC matters

Mark Scholten mark at streamservice.nl
Tue Sep 4 02:12:35 UTC 2012

Hello Peter,

I removed some parts from your email to make it easier to read.

> ISSUE 1: ordername sorting
> 1a. How do we tell Postgres to do "the right thing" for us, preferably
> in a way that does not force all users to do a dump/restore? We
> wouldn't mind an ALTER TABLE or the like!
> 1b. Is VARBINARY the best way to do it for MySQL?
> 1c. Should we cave in and encode ordername in some way that will sort
> reliably, regardless of database settings? Base64 perhaps? This does
> involve stretching ordername beyond 255 chars, which presumably is okay
> with all common versions of PG, My and SQLite3.

1a. I don't use Postgres and I don't know enough about it to say anything
about it.
1b. When it works and is fast it would sound like a good option for me.
1c. Encoding it also requires that it is decoded when it is requested? That
sounds like extra processor/RAM usage and I'm not a fan of that.

> ISSUE 2: non-empty terminals
> 2a. Do you think adding these records to the records table is sensible
> at all? If not, how else would we do it?
> 2b. Do you think type=NULL (SQL NULL) is an ugly hack? If so, what else
> should we do?
> 2c. If we accept type=NULL as an acceptable notation, should we still
> have this extra field just to make cleanup easier?

2a. It sounds like they are missing, so adding them sounds like the best
2b. Type=pdns sounds better to me (that will probably never be used in an
RFC and makes it clear when reading the database that it is an internal
record for PowerDNS.
2c. If type= says that it is automatically created/maintained I don't think
the extra field will be necessary (at least when the right indexes are
created/are there).

Regards, Mark

More information about the Pdns-users mailing list