[Postgres-xl-developers] synchronous replica

Mason Sharp msharp at translattice.com
Mon Aug 25 10:50:43 PDT 2014


Hi Bryain,

Yes, I was saying that if you use pgxc_ctl with the latest code, you can
save yourself from having to work through so many of these issues. Please
try that.

Thanks,

Mason



On Sun, Aug 24, 2014 at 11:15 PM, bryain <pgdevuser at 163.com> wrote:

> Hi,
> Thank you for your patience reply.
> There is a configuration problem.We need not only configure the
> coordinator node, also the datanode node,then solved.
>
> Bryain
>
>
> 在 2014-08-22 09:01:41,"Mason Sharp" <msharp at translattice.com> 写道:
>
>
>
>
> On Fri, Aug 22, 2014 at 12:57 AM, bryain <pgdevuser at 163.com> wrote:
>
>> Hi Mason Sharp,
>> I finished the cluster configuration.
>> the database cluster with 1 gtm, 2 gtm_proxy,2 coordinates and 2
>> datanodes(GTM in a separate physical server,and others 2 in a independent
>> of a physical server).
>>
>> exec "select * from pgxc_node" return 4 rows as blow:
>>  node_name | node_type | node_port | node_host | nodeis_primary |
>> nodeis_preferred |   node_id
>>
>> -----------+-----------+-----------+-----------+----------------+------------------+-------------
>>  node1     | D         |     15432 | node1     | f              |
>> f                |   888802358
>>  node2     | D         |     15432 | node2     | f              |
>> f                |  -905831925
>>  coord1    | C         |      5432 | node1     | f              |
>> f                |  1885696643
>>  coord2    | C         |      5432 | node2     | f              |
>> f                | -1197102633
>>
>> *when I create database,*
>> postgres=# create database test_xl;
>> return error msg:
>> LOG:  Will fall back to local snapshot for XID = 32138, source = 0, gxmin
>> = 0, autovac launch = 0, autovac = 0, normProcMode = 0, postEnv = 1
>> ERROR:  node "coord1_9679" does not exist
>> STATEMENT:  SET global_session TO coord1_9679;
>> ERROR:  node "coord1_9679" does not exist
>> STATEMENT:  SET global_session TO coord1_9679;
>> CREATE DATABASE
>>
>> *atfer that,The database has been created.*
>> *create table for testing,*
>> create table test1(id integer,name varchar(20));
>> CREATE TABLE
>> insert into test1 values(1,'a');
>> ERROR:  node "coord1_9755" does not exist
>> STATEMENT:  SET global_session TO coord1_9755;
>> INSERT 0 0
>> there is no data have be inserted.
>>
>> *Bulk insert data*
>> test=# insert into test1 select generate_series(1,100),'test_xl';
>> ERROR:  node "coord1_9949" does not exist
>> STATEMENT:  SET global_session TO coord1_9949;
>> ERROR:  Invalid Datanode number
>> STATEMENT:  Remote Subplan
>> LOG:  Connection error Invalid Datanode number
>> STATEMENT:  insert into test1 select generate_series(1,100),'test_xl';
>> LOG:  Connection error Invalid Datanode number
>> STATEMENT:  insert into test1 select generate_series(1,100),'test_xl';
>> ERROR:  Invalid Datanode number
>> STATEMENT:  insert into test1 select generate_series(1,100),'test_xl';
>> ERROR:  Invalid Datanode number
>>
>> Where is the usually reasons for this problem?thanks!
>>
>>
> Bryain,
>
> Do you mind starting over and using the pgxc_ctl utility? It is much
> easier to build up a working cluster that way than to track down what
> particular step was missed in the configuration. You will end up saving
> yourself time in the long run.
>
> I would get a configuration working without GTM proxy, then use pgxc_ctl
> to add a GTM Proxy where each Coordinator is running.
>
> Also, please run with the latest code in the git repository at
> sourceforge, it contains bug fixes for pgxc_ctl.
>
> Regards,
>
> --
> Mason Sharp
>
> TransLattice - http://www.translattice.com
> TransLattice Elastic Database (TED)
> Postgres-XL Support
>
>
>
>


-- 
Mason Sharp

TransLattice - http://www.translattice.com
TransLattice Elastic Database (TED)
Postgres-XL Support
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.postgres-xl.org/private.cgi/postgres-xl-developers-postgres-xl.org/attachments/20140825/ce7b5a10/attachment.htm>


More information about the Postgres-xl-developers mailing list