BIND 10 #2862: update b10-auth to recognize data source memory segments
BIND 10 Development
do-not-reply at isc.org
Fri Jul 12 09:52:34 UTC 2013
#2862: update b10-auth to recognize data source memory segments
-------------------------------------+-------------------------------------
Reporter: jinmei | Owner:
Type: task | UnAssigned
Priority: medium | Status:
Component: b10-auth | assigned
Keywords: | Milestone:
Sensitive: 0 | Sprint-20130723
Sub-Project: DNS | Resolution:
Estimated Difficulty: 5 | CVSS Scoring:
Total Hours: 0 | Defect Severity: N/A
| Feature Depending on Ticket:
| shared memory data source
| Add Hours to Ticket: 0
| Internal?: 0
-------------------------------------+-------------------------------------
Changes (by vorner):
* owner: vorner => UnAssigned
* status: accepted => assigned
Comment:
Hello
It is ready for review. However, there are some points worth discussing.
There are some new functionalities in the `ModuleCCSession`. It seems we
are bending it for something that it doesn't want to do right now
(pretending to be some other module). Also, part of that new functionality
is not really clean. I needed something, but the ticket is taking a long
time already and it didn't seem directly relevant, so I hacked it in. But
a real solution would be needed.
Also, there's a related part of the auth code that is not tested, because
it sounds it is very hard to test (see the last commit). I think this will
be better tested by a lettuce test (and that we'll need one in the end
anyway, because I'm bit worried if the communication in auth will match
the one in the memmgr).
So, there are two possibilities. Create a ticket to clean up the
`ModuleCCSession`, check that there's a ticket for the lettuce test. Or
the other is to redesign the communication with primitives we already have
(eg. notifications and commands).
The other thing is, what to do if the update of the segment fails. We
currently just skip it (hoping it is just out of sync with the memmgr and
because of different configuration, and that'll will get synced soon). But
we ACK it while still using the old segment (possibly). Should we release
the segment? Or detect the error somehow and signal it?
--
Ticket URL: <http://bind10.isc.org/ticket/2862#comment:6>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list