[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: NETCONF Notifications: Consensus Points
To review, the times when one needs an "application level
acknolwedgment" (ALA) is when the data has received and acknowledged by
the sender, and further processing then fails. The classic example is a
disk failure.
Well, what comes next? If further processing fails does the receiver
remain up or can it tear down the connection if its configuration
dictates? If not, then ALAs are needed.
If so, the other end has a signal that there is in fact a problem, and
that it should follow its policy (for instance, going to backup server,
re-establishing connection, what-have-you). At that point, if
notifications are general in nature (e.g., not related to configuration,
then one could argue that unless alarms are part of the architecture,
ALAs will be needed because traipsing through device state searching for
faults is probably cost prohibitive.
If the notifications are related to configuration alone, will the new
notification recipient be able to retrieve sufficient state to determine
a configuration change in the device? If the answer is yes, then no
ALAs are needed.
Right or wrong?
Eliot
--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>