http://www.slony.info/bugzilla/show_bug.cgi?id=235

Steve Singer <ssinger at ca.afilias.info> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|ssinger at ca.afilias.info     |cbbrowne at ca.afilias.info

--- Comment #13 from Steve Singer <ssinger at ca.afilias.info> 2011-11-24 07:10:09 PST ---
I've taken a lot at the rebased against the current master (2.2).


--- 249,263 ----
  time_t        explain_lastsec;
  int            explain_thistime;

! typedef enum
! {
!     SYNC_INITIAL = 1,
!     SYNC_PENDING,
!     SYNC_SUCCESS,
!     SYNC_FAILURE
! 
! } SlonSyncStatus;

SYNC_FAILURE is a state that is never actually set/checked.  In the event of a
failure applying the sync the remote worker causes the slon to restart
resetting the count to the inital state.  None of the logic to decrease the
group size is used.   We should either get the remote_worker to detect+recover
from this error OR remove the FAILURE state and /2 logic and add a comment that
we depend on restart.

In confoptions.c  
{
        {
            (const char *) "sync_group_maxsize",
            gettext_noop("sync group"),
            gettext_noop("sync group"),
            SLON_C_INT
        },
        &sync_group_maxsize,
        20,
        0,
        10000
    },

isn't part of you diff but should be.   The description (inside gettext_noop)
is insufficient.  It should say something like "The maximum number of SYNC
events that will be applied as part of the same transaction"   

I tested the doubling logic and it seemed fine.

-- 
Configure bugmail: http://www.slony.info/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.