On 9/17/2004 2:01 PM, Ed L. wrote:

> A few questions on slon command-line argument semantics...
> 
> 
> % slon -h
> usage: slon [options] clustername conninfo
> 
> Options:
>     -d <debuglevel>       verbosity of logging (1..4)
>     -s <milliseconds>     SYNC check interval (default 10000)
>     -t <milliseconds>     SYNC interval timeout (default 60000)
>     -g <num>              maximum SYNC group size (default 6)
>     -c <num>               how often to vaccum in cleanup cycles
> 
> I'm assuming "-s" means check for updates needing propagation this often?

Right.

> 
> What does the ("-t") SYNC interval timeout mean?  Is that how long a sync is 
> allowed to progress before being aborted?  How long to wait on checking a 
> slave's status??

It means that it will a SYNC event at least that often, regardless of 
detected update activity or not. The problem here is a possible race 
condition. If the action sequence is bumped by the trigger while 
inserting the log row, this bump is immediately visible to the sync 
thread. But the resulting log rows are not visible yet, so if the sync 
is picked up by the subscriber, processed and finished before the 
transaction commits, this transactions changes will not get replicated 
until the next SYNC. But if now out of the blue all application activity 
stops, there will be no more sequence bumps, so the high frequent -s 
check won't detect that.

> 
> What are the units for the -g flag?  Transactions?  Rows?

Number of SYNC events to cram together.

> 
> How do I disable slon-initiated vacuuming altogether?  (Our vacuuming is 
> handled by an autovac.)

-c 0


Jan

-- 
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck at Yahoo.com #