Ok - I will check the sync then. Thanks. On Fri, Apr 6, 2018 at 7:44 PM David Christensen <[email protected]> wrote:
> > > On Apr 6, 2018, at 3:10 AM, Raghu Udiyar <[email protected]> wrote: > > Hi > > We are seeing that row deletes are not being propagated to the target PG > db. This is a PG one way sync using Bucardo 5.4.1 > (using quick_delta_check=0). > > In Bucardo logs we always see *deletes=0*, but deletes are happening on > the table in question. > > I was trying to figure out how deletes are handled in Bucardo. I can see > the DELETE trigger adds the OLD values to the corresponding delta table, > but the table does not have any field to indicate if this is a DELETE or > not. The same is done for INSERT albeit using the NEW values. Only thing > added is the *txntime*. Is this used to implicitly delete records as a > side effect of an row insert/update ? > > Thanks > Raghu > > > Hi Raghu, > > DELETEs are noted as a delta row with no corresponding row in the table; > basically any row which is modified stores the PK and the transaction > time. INSERTs/UPDATEs are handled the same way; effectively Bucardo will > delete any keys on the target with PKs in the delta tables, then copy any > data from the original table with the same keys. In the case of DELETEs, > those will not exist, so the row is effectively deleted. > > Have you verified that INSERTs and UPDATEs are working on the target? I’m > a little curious why it would just affect DELETEs and not the other types, > so I’d first verify that the sync is working at all. > > Best, > > David > -- > David Christensen > End Point Corporation > [email protected] > 785-727-1171 <(785)%20727-1171> > > > > -- -- [image: photo] <https://helpshift.com/> Raghu Udiyar Manager - Production Engineering [image: https://in.linkedin.com/in/parul-jain-7b489830] <https://www.linkedin.com/in/raghusiddarth/>
_______________________________________________ Bucardo-general mailing list [email protected] https://mail.endcrypt.com/mailman/listinfo/bucardo-general
