BIND 10 #2380: revise b10-loadzone using datasrc.ZoneLoader
BIND 10 Development
do-not-reply at isc.org
Tue Dec 18 13:08:23 UTC 2012
#2380: revise b10-loadzone using datasrc.ZoneLoader
-------------------------------------+-------------------------------------
Reporter: jinmei | Owner: jelte
Type: task | Status:
Priority: medium | reviewing
Component: loadzone | Milestone:
Keywords: | Sprint-20121218
Sensitive: 0 | Resolution:
Sub-Project: DNS | CVSS Scoring:
Estimated Difficulty: 8 | Defect Severity: N/A
Total Hours: 0 | Feature Depending on Ticket:
| loadzone-ng
| Add Hours to Ticket: 0
| Internal?: 0
-------------------------------------+-------------------------------------
Comment (by jelte):
Replying to [comment:10 jinmei]:
>
> For the temporary measure for the beta release, I propose this: allow
> omitting -c for sqlite3 data source (which is the only supported data
> source at the moment anyway). When omitted, it internally uses the
> configuration with specifying the default sqlite3 DB file. It's still
> datasrc-specific hack hardcoded in the source, but at least the user
> interface is still datasrc-agnostic.
>
> I've implemented this proposed approach.
>
OK this seems a reasonable compromise
> I've tried to improve it a bit. But that's not a perfect as you
> noticed, of course. We should create a ticket for getting the total
> number of loaded RRs and use it. We should also be able to print
> the progress in percentage. These require extensions to the
> underlying API, and will consist of a few tickets.
>
Yeah, either a getNrOfLoadedRRs (with a better name :p) or maybe change
the return value of loadIncremental to nr-of-rrs-loaded (if < limit,
you're done), though that would be a backwards incompatible change
>
> > Tests fail for me (I build with --prefix but without installing):
> >
> > I guess the Makefile needs to set B10_FROM_BUILD to get the correct
dynamic loading dir.
>
> Ah, right, added.
>
I ran into another test issue; the 'rename-process' test flags b10
-loadzone-ng as bad, it either needs to whitelist it or -ng needs to call
the rename process function like the original (but i think we can just
whitelist it)
>
> > maybe a matter of taste, but should we not consider a zone without a
SOA a complete failure and error/revert on it?
>
> We should, but I'd like to defer it to the forcecoming validator (that
> will run under the updater, so the rollback will automatically happen
> once we implement validator and integrate it in the zone loader). We
> could make this cleanup as a temporary hack here, but that would be
> much more complicated than just canceling adding the zone, so I opted
> to not do too much here.
Fair enough.
--
Ticket URL: <http://bind10.isc.org/ticket/2380#comment:12>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list