[bind10-dev] Statistics planning

[JINMEI Tatuya / 神明達哉]

At Mon, 21 Nov 2011 18:35:02 +0900 (JST),
Naoki Kambe <kambe at jprs.co.jp> wrote:

> As broken-down statistics tasks in next two sprints, we, JPRS
> members planned as the following.
> 
>   2011-11-23 - 2011-12-06
>    - Implement dictionary-type variable size counter (#510)
>    - implement a prototype for stats snmp interface (#916)
>   2011-12-07 - 2011-12-20
>    - add some statistics itmes to auth after #510 (#TBD1)
>    - document basic design of stats snmp interface (#TBD2)
> 
> In the next print, we're thinking of #510 and #916. #510 is taken over
> from the last sprint by Aharen. #916 was in "Year 3 Task Backlog", I
> already moved it into Next-Sprint-Proposed. After that next print,
> Aharen will add some statistics itmes to auth after #510 (#TBD1) and I
> will document design of stats snmp interface (#TBD2). That is our
> proposal of statistics planning in the near future. If it isn't ok for
> you, please let me know.

Nothing is not okay for me, as I'm basically just making suggestions,
not (dis)approving your plans:-)

My question/comment about the above plans are:
- how likely will we complete #510 in the next sprint?  It's given an
  estimation of 3, yet has been open for about 4 months, and has been
  in the review queue for about 3 months.  My suggestion is that if we
  still cannot close this in the next sprint, possibly because
  Aharen-san is too busy, we'll reassign it to someone else in the
  sprint starting from 2011-12-07 so that we can at least make modest
  progress.
- what do you think about my higher level proposal that we include at
  least some user visible feature regarding statistics in every
  snapshot release?  If you are positive about it, do you think it
  makes sense to set it as the goal for the next release to support a
  substantially large number of more statistics auth counters?
- what do you think about my suggestion on how others than JPRS
  developers are involved in statistics related development?  The
  relevant part of my previous comments are:
  - As for the task sizes, IMO other developers than JPRS should be
    more involved in task breakdown at the sprint planning meeting,
    or, preferably, in a pre-meeting phase for the planning meeting
    using email.  I believe others should be involved because it's
    them (us) who will eventually review the code even if the JPRS
    developers solely work on the tasks.
  - Also, in terms of releasing a specific statistics feature, I
    believe we (all) should even be more closely involved in the
    development process.  For example, we keep track of the progress
    of statistics for the first two sprints in a release cycle, and if
    it's less likely that we can complete the visible feature by the
    release we'll include the remaining tasks in other planned tasks
    at a higher priority and everyone will work on it.
  (The specific suggestion on #510 in the first bullet is one example
  of the closer involvement)

I'd like to discuss these as part of the sprint planning meeting
tomorrow (my time).

> BTW we can see other new statistics tickets which are already opened
> by Jinmei-san, those are fine to us.

Okay, good, then please don't forget give estimates to them and send
the results to Jelte (if you've not done that:-).

---
JINMEI, Tatuya
Internet Systems Consortium, Inc.