Separating Peering from PG

[Loic Dachary <loic@dachary.org>]

[-- Attachment #1: Type: text/plain, Size: 980 bytes --]

Hi Sage,

In order to move the PG peering code out of PG.{cc,h} (which is the next step in refactoring PGs as suggested by Sam http://pad.ceph.com/p/Erasure_encoding_as_a_storage_backend ) I think it would be sensible to:

* Move PG::RecoveryStats in PGRecoveryStat.{cc,h}
* Create PGInterface : an abstract base class for PG enumerating all PG methods used by PGRecoveryStats
* Move Peering states / methods out of PGRecoveryStat.{cc,h} and into PGPeering.{cc,h} 
* Write tests for PGPeering.{cc,h}, using a fixture derived from PGInterface

Because this approach not only moves the peering out of PG.{cc,h} but also the rest of the state logic, I would like to know if this seems sensible to you. Also, introducing an abstract base class to help isolate the PG interface and facilitate writing fixtures has not been discussed yet. 

Cheers

-- 
Loïc Dachary, Artisan Logiciel Libre
All that is necessary for the triumph of evil is that good people do nothing.


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]