canonical-ci-engineering team mailing list archive

[Vincent Ladeuil <vila+ci@xxxxxxxxxxxxx>]

>>>>> Evan Dandrea <evan.dandrea@xxxxxxxxxxxxx> writes:

    > I don't know of a case where we're using the nack mechanism for
    > anything else. Can you provide an example, and how you'd rather it be
    > implemented?

No we don't use but I keep hearing about using it (no later than during
a discussion with pitti (who also couldn't found a way to
properly/elegantly do that with rabbit)).

The use case is around britney issuing a test request received by a
worker that cannot fulfill it (in a context where some queues can
receive requests for different arches or hardware constraints).

In that case, instead of nacking the message hoping that some other
worker will be able to handle it, I argued that:

1) as a rule, workers should only receive requests they can handle,

2) if they fail to fulfill the request (or find they *cannot* process
   the request) they should reply with an error message and not put the
   message back on the queue

Sorry for the confusion,

      Vincent