Thread: about postgres-r setup.

hi,
i have setup a postgres-r environment, including two nodes, and
1) the Ensemble daemon run with default configure (means no command options)
2) set the internal variable
repl_default_gcs = 'ensemble'
at postgresql.conf
3) all initial sql script initialized at node1 and node2 as follows:
create database data;
\c data;
create table foo (id serial primary key, n text);
4) execute sql:
alter database data start replication in group tst;

at node1, and seems everything is ok:

NOTICE: Applying for membership in the communication group...
NOTICE: Successfully started replication.
ALTER DATABASE

5)execute sql:
alter database data accept replication from group tst;

at node2, and the psql command is hung up,

NOTICE: Applying for membership in the communication group...
NOTICE: Successfully joined the replication group. Now recovering schema...

so, i dont know what happened, pls give me some advice.

thanks.

best,
leiyonghua

Hello Leiyonghua,

thanks for trying out Postgres-R.

leiyonghua wrote:
> NOTICE: Applying for membership in the communication group...
> NOTICE: Successfully joined the replication group. Now recovering schema...
> 
> so, i dont know what happened, pls give me some advice.

Please compile with '--enable-debug' and '--enable-cassert' and check 
the log.

The schema recovery has not been implemented, yet. So I find it pretty 
suspicious it is hanging there. Although, I must admit I didn't try the 
ensemble interface layer for a while.

Regards

Markus

Hello Leiyonghua,

leiyonghua wrote:
> at node2, and the psql command is hung up,
> 
> NOTICE: Applying for membership in the communication group...
> NOTICE: Successfully joined the replication group. Now recovering schema...

I've just tested the very same here with ensemble, and could reproduce 
the situation. It had to do with the ensemble receive buffer being too 
small to hold the recovery data. That's fixed in today's snapshot. With 
that I've been able to replicate simple tuples again.

Regards

Markus Wanner

Markus Wanner 写道:
> Hello Leiyonghua,
>
> leiyonghua wrote:
>> at node2, and the psql command is hung up,
>>
>> NOTICE: Applying for membership in the communication group...
>> NOTICE: Successfully joined the replication group. Now recovering 
>> schema...
>
> I've just tested the very same here with ensemble, and could reproduce 
> the situation. It had to do with the ensemble receive buffer being too 
> small to hold the recovery data. That's fixed in today's snapshot. 
> With that I've been able to replicate simple tuples again.
>
that's cool !  i will try later. 
> Regards
>
> Markus Wanner
>
>
>
leiyonghua

leiyonghua 写道:
> Markus Wanner 写道:
>> Hello Leiyonghua,
>>
>> leiyonghua wrote:
>>> at node2, and the psql command is hung up,
>>>
>>> NOTICE: Applying for membership in the communication group...
>>> NOTICE: Successfully joined the replication group. Now recovering 
>>> schema...
>>
>> I've just tested the very same here with ensemble, and could 
>> reproduce the situation. It had to do with the ensemble receive 
>> buffer being too small to hold the recovery data. That's fixed in 
>> today's snapshot. With that I've been able to replicate simple tuples 
>> again.
>>
> that's cool !  i will try later.
>> Regards
>>
>> Markus Wanner
>>
>>
>>
> leiyonghua
>
today, i try the snap-shot

postgres-r-20080813.diff

and still same status. 

leiyonghua

Hi leiyonghua,

leiyonghua wrote:
> and still same status.

Uh.. do you have debugging enabled? Any logging output of the two 
postmaster processes?

Regards

Markus