Ricardo Manuel Pereira Vilaça schrieb:
> Hi Ewald
>> Thanks - we try to do our best and if we can we contribute to the
>> project - maybe put more workforce in it.
>>
>> One question can you add nodes to the cluster or can you make catchup
>> of some nodes?
>>
> Yes. The replication engine use on-line recovery and synchronize
> recovery with normal processing.
> Currently we use the PostgreSQL tools for this purpose, pg_dump and
> pg_restore.
> We are working on other recovery protocols that just transfer the
> missed updates.
>
thanks for the info we have implemented a multimaster in java for pg one
year ago - but it was too slow
and we have solved the problem for this missed updates with an
in-memory queue.
When is the expected release of the rep server - I'm hungry for testing ;)
Regards Ewald Geschwinde
