[Ppm] Adding support to DAP for fixed-chunk tasks

[Christopher Patton <[e-mail]>]

Hi all,

I've created a PR [1] to add support for the use case described in [2]. The
PR Is designed to fulfill the following requirements, which were discussed
in the PPM meeting at IETF 114. (Refer to the slides for the use cases
presentation [3].)

1. Enumerates all query types and defines the behavior of the Client,
Leader, Helper, and Collector for each.
2. Two query types are defined, one that supports the "time-series
telemetry" use case envisioned by the current draft. The other supports the
use case discussed in [2], which requires the collector to be able to
strictly control the batch size.
3. Punts on the "drill down by client property" use case discussed in [4].
We decided in the meeting that we have not yet fleshed out the requirements
enough to warrant changes at this time.
4. Allows both Aggregators to ensure that each report in a batch satisfies
the query.
5. Allows implementations to add new query types as needed.
6. Permits only one query type per task, which allows us to punt on the
complexity of composing different query types. The query type is configured
out-of-band, along with the rest of the task configuration.

If folks have questions or concerns about these requirements, I think it
would be best to discuss them here rather than on the PR. Otherwise, review
of the PR itself is greatly appreciated!

Best,
Chris P.

[1] https://github.com/ietf-wg-ppm/draft-ietf-ppm-dap/pull/297
[2] https://github.com/ietf-wg-ppm/draft-ietf-ppm-dap/issues/273
[3]
https://datatracker.ietf.org/meeting/114/materials/slides-114-ppm-dap-use-cases-00
[4] https://github.com/ietf-wg-ppm/draft-ietf-ppm-dap/issues/183