Geoffrey wrote:
> Cyril SCETBON wrote:
>> Which type of ALTER did you make ?
>>
>> Don't forget that there's a special trigger on the source table that 
>> must be updated too.
>
> I'm not sure of how the change was made, I didn't make it.  As far as 
> I know, all he did was add a new field to one of the tables.  He did 
> not modify any slony triggers by hand.
Not good. Check in the slon daemons logs to see what's happening.
>
>>
>> Geoffrey wrote:
>>> David Rees wrote:
>>>> On Tue, Nov 4, 2008 at 2:22 PM, Geoffrey 
>>>> <lists at serioustechnology.com> wrote:
>>>>> Will it cause a problem if one adds fields to a table that is being
>>>>> replicated?
>>>>>
>>>>> That is, will it break the replication process, or will the 
>>>>> replication
>>>>> continue to simply replicate the previous set of fields?
>>>>
>>>> It should continue to replicate the previous set of fields, but is not
>>>> recommended to do so.
>>>
>>> I don't know if it's related or not, but it appears that our slave 
>>> is now out of sync with the master.  It's a two node set up, and 
>>> it's been replicating for some time now.  It's just now that I've 
>>> gone in and compared record counts on some major tables and they are 
>>> not in sync. Further, I can see updates happen to the master node 
>>> tables but don't see them on the slave.
>>>
>>> If adding the field didn't cause the problem, any ideas what might 
>>> have?
>>>
>>
>
>