BIND 10 #3360: memfile: implement file support
BIND 10 Development
do-not-reply at isc.org
Thu Apr 3 09:49:22 UTC 2014
#3360: memfile: implement file support
-------------------------------------+-------------------------------------
Reporter: tomek | Owner:
Type: enhancement | marcin
Priority: medium | Status:
Component: dhcpdb | reviewing
Keywords: | Milestone: DHCP-
Sensitive: 0 | Kea0.9-alpha
Sub-Project: DHCP | Resolution:
Estimated Difficulty: 5 | CVSS Scoring:
Total Hours: 2 | Defect Severity: N/A
| Feature Depending on Ticket:
| Add Hours to Ticket: 2
| Internal?: 0
-------------------------------------+-------------------------------------
Changes (by stephen):
* owner: stephen => marcin
Comment:
Reviewed commit d78c00959f21b45797d20bf3775688ae163528e9
> I want to stick to the concept of the default file location. The default
is taken when the file is not specified or is empty. This is not really
doable if we want to use null or not null to disable or enable
persistence. This can't have two meanings.
As the "type" keyword already has a default value of "memfile", the same
mechanism could provide a default value for "name" after installation,
that is up to them.
> I don't want to remove the file name from the configuration to
temporarily disable persistence because I will have hard time recreating
the name when I want to re-enable it. It is so much easier for the user to
flip boolean flag, rather than copy back and forth the absolute file path.
It is unlikely that users will flip back and forth between running with a
lease file and not running with one. So I don't think this really an
issue, although I can see it could be useful in some circumstances.
What I'm most worried about with the "persist" flag is a user
inadvertently setting "persist" to false but, seeing that "name" is set
and that there is a file of that name on disk, believing that the system
is persisting leases when it isn't.
However, this is something we can alter later if need be. Leave it as it
is for the 0.8 release.
> Finally, is .spec file really an issue? We are going to toss our
configuration mechanism quite soon anyway. I also remember that defaults
didn't really work properly through the spec file.
Probably not.
All OK, please merge.
--
Ticket URL: <http://bind10.isc.org/ticket/3360#comment:8>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list