postgres-xc-developers Mailing List for Postgres-XC (Page 138)

On 10/14/10 3:30 AM, mei le wrote:
> Dears,
> Postgres-xc version 0.9.2 has error process at the end of funcion 
> "ExecRemoteQuery" but such a part was deleted by later patch.In 
> postgres-xc 0.9.2, if I create a table on one datanode, I can't create 
> table on other datanodes by executing create table statment on 
> Coordinator, but the newest code can create table on other datanodes 
> by executiong same statement on Coordinator.I think the former version 
> is more appropriate than the newest code. Could you give me some 
> explains on this issue?
> Thanks.
>                                                                                       
> Truely,
>                                                                                              
> Benny
>
>

Benny,

I am sorry, can you please explain again, I think I do not understand.

In Postgres-XC, when you create a table, you do it against a 
Coordinator, and not a Data Node. The Coordinator will automatically 
create the table on the Data Nodes for you. It should be this way for 
both 0.9.2 and the latest code in the git repository.

Please let me know if I misunderstood something.

Thanks,

Mason


>
> ------------------------------------------------------------------------------
> Beautiful is writing same markup. Internet Explorer 9 supports
> standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2&  L3.
> Spend less time writing and  rewriting code and more time creating great
> experiences on the web. Be a part of the beta today.
> http://p.sf.net/sfu/beautyoftheweb
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>    


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Dears,
 
Postgres-xc version 0.9.2 has error process at the end of funcion "ExecRemoteQuery" but such a part was deleted by later patch.In postgres-xc 0.9.2, if I create a table on one datanode, I can't create table on other datanodes by executing create table statment on Coordinator, but the newest code can create table on other datanodes by executiong same statement on Coordinator.I think the former version is more appropriate than the newest code. Could you give me some explains on this issue?
 
Thanks.
 
                                                                                      Truely,
                                                                                             Benny


      

Thanks, I will look at it soon more carefully.

A manual checkpoint should not occur too often, so I am not that 
concerned about efficiency for this case. Can the other code you did be 
used for other cases, too?

Thanks,

Mason


On 10/13/10 5:35 AM, mei le wrote:
> Dears,
> I made two patches for bug 
> https://sourceforge.net/tracker/?limit=25&func=&group_id=311227&atid=1310232&assignee=&status=&category=&artgroup=&keyword=&submitter=&artifact_id=&assignee=&status=1&category=&artgroup=&submitter=&keyword=&artifact_id=&submit=Filter&mass_category=&mass_priority=&mass_resolution=&mass_assignee=&mass_artgroup=&mass_status=&mass_cannedresponse= 
> <https://sourceforge.net/tracker/?limit=25&func=&group_id=311227&atid=1310232&assignee=&status=&category=&artgroup=&keyword=&submitter=&artifact_id=&assignee=&status=1&category=&artgroup=&submitter=&keyword=&artifact_id=&submit=Filter&mass_category=&mass_priority=&mass_resolution=&mass_assignee=&mass_artgroup=&mass_status=&mass_cannedresponse=>
> 3076224    checkpoint command causes seg fault
> 3010262    Primary Key in CREATE TABLE
> 3013984    Sequence scope
> 3013562    Multiple insert not possible
> The first one is the most simple method to solve the bug, but it may 
> lower efficency. So, I made second one which did special process for 
> checkpoint which doesn't need transaction process.
> Looking forward your kindly reply.
>                                                                                       
> Truely,
>                                                                                               
> Benny
> --- *10年9月30日，周四, Mason Sharp /<mas...@en...>/* 
> 写道：
>
>
>     发件人: Mason Sharp <mas...@en...>
>     主题: tasks
>     收件人: "mei le" <lem...@ya...>
>     抄送: ""03：鈴木(幸)様(DI)"" <ko...@in...>
>     日期: 2010年9月30日,周四,下午4:46
>
>     I understand you and your friends are busy for a while.
>
>     If you would like to get involved, once you have more time later,
>     here are some tasks that might be some good ones to start with, in
>     approximate order of what makes sense to work on.
>
>
>
>     Bugs
>
>     https://sourceforge.net/tracker/?limit=25&func=&group_id=311227&atid=1310232&assignee=&status=&category=&artgroup=&keyword=&submitter=&artifact_id=&assignee=&status=1&category=&artgroup=&submitter=&keyword=&artifact_id=&submit=Filter&mass_category=&mass_priority=&mass_resolution=&mass_assignee=&mass_artgroup=&mass_status=&mass_cannedresponse=
>     <https://sourceforge.net/tracker/?limit=25&func=&group_id=311227&atid=1310232&assignee=&status=&category=&artgroup=&keyword=&submitter=&artifact_id=&assignee=&status=1&category=&artgroup=&submitter=&keyword=&artifact_id=&submit=Filter&mass_category=&mass_priority=&mass_resolution=&mass_assignee=&mass_artgroup=&mass_status=&mass_cannedresponse=>
>
>     3076224    checkpoint command causes seg fault
>     3010262    Primary Key in CREATE TABLE
>     3013984    Sequence scope
>     3013562    Multiple insert not possible
>
>
>
>     Features
>
>     https://sourceforge.net/tracker/?group_id=311227&atid=1310235
>     <https://sourceforge.net/tracker/?group_id=311227&atid=1310235>
>
>     3000104    COPY TO with SELECT
>     3014003    Pooler Request Queuing
>     3014002    Pooler should shrink
>     3022406    Safer Hash Distribution to Data Nodes
>
>
>     Thanks,
>
>     Mason
>
>     -- Mason Sharp
>     EnterpriseDB Corporation
>     The Enterprise Postgres Company
>
>
>     This e-mail message (and any attachment) is intended for the use of
>     the individual or entity to whom it is addressed. This message
>     contains information from EnterpriseDB Corporation that may be
>     privileged, confidential, or exempt from disclosure under applicable
>     law. If you are not the intended recipient or authorized to receive
>     this for the intended recipient, any use, dissemination, distribution,
>     retention, archiving, or copying of this communication is strictly
>     prohibited. If you have received this e-mail in error, please notify
>     the sender immediately by reply e-mail and delete this message.
>
>


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Dears，
 
When I read code，I found something wrong in fuction "commit_prepared_transaction" which maybe a mistake. I think there should be one sentence that "return 0;" before label "send_failed" in function "commit_prepared_transaction".
 
                                                                        Truely,
                                                                           Benny 


      

When I use checkpoint command, the coordinator report the error as belows.Is it a bug or something?
 
template1=# checkpoint;
server closed the connection unexpectedly
 This probably means the server terminated abnormally
 before or while processing the request.
The connection to the server was lost. Attempting reset: LOG:  server process (PID 20650) was terminated by signal 11: Segmentation fault
LOG:  terminating any other active server processes
LOG:  unexpected EOF on client connection
LOG:  all server processes terminated; reinitializing
LOG:  database system was interrupted; last known up at 2010-09-26 11:41:16 CST
LOG:  database system was not properly shut down; automatic recovery in progress
FATAL:  failed to connect to pool manager: Connection refused
Failed.
!> LOG:  record with zero length at 0/4D9538
LOG:  redo is not required


      

Benny,

An additional comment-

In the source code, most of the changes compared to PostgreSQL are below 
src/backend/pgxc. These are all new files. You can look through and read 
comments.

In addition, changes to existing PostgreSQL modules can be easily found 
by searching for the string PGXC, so you can see what changed and why.

Thanks,

Mason


On 9/20/10 12:15 AM, Koichi Suzuki wrote:
> Benny,
>
> Thank you very much for the suggestion.    Yes, I understand the 
> README is not rich.   Instead, we have "architecture" document which 
> you can download separately, as well as other manual and guidelines.
>
> Could you try these documents?   Especially, I made "architecture" 
> document as a separate one because XC internal is not simple to 
> understand and I thought usual document with graphics and tables will 
> be more helpful.
>
> Your advice is very valuable.   I'll be waiting for your further advice.
>
> Thank you very much;
> ----------
> Koichi Suzuki
>
>
> 2010/9/19 mei le <lem...@ya... 
> <mailto:lem...@ya...>>
>
>     Dears :
>
>     This is Benny ,a postgraduate for DBMS from DUTchina. I have
>     studied postgreSQL for a long time. And I have always kept close
>     eye on postgreSQL cluster solutions such as PGCluster-2 and
>     GridSQL etc. After postgres-xc released, I have learned much about
>     it. It's a great cluster frame. I have been reading postgres-xc
>     source code recently.
>
>              Reading PostgresSQL code, I always like to read README
>     under relative directory, which can make me learn more knowledge
>     on its relative functions. But when I scan postgres-xc code which
>     was freshly added, I can't find any README to introduce the
>     summary or relative knowledge. I think README is very important to
>     an open source code. I don't know why postgres-xc ignores README
>     files .Could we add the README files later?
>
>                                                                                              
>     Truly yours,
>
>     Benny
>
>
>
>     ------------------------------------------------------------------------------
>     Start uncovering the many advantages of virtual appliances
>     and start using them to simplify application deployment and
>     accelerate your shift to cloud computing.
>     http://p.sf.net/sfu/novell-sfdev2dev
>     _______________________________________________
>     Postgres-xc-developers mailing list
>     Pos...@li...
>     <mailto:Pos...@li...>
>     https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>
>
>
> ------------------------------------------------------------------------------
> Start uncovering the many advantages of virtual appliances
> and start using them to simplify application deployment and
> accelerate your shift to cloud computing.
> http://p.sf.net/sfu/novell-sfdev2dev
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>    


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

On 9/20/10 4:19 AM, mei le wrote:
> Dears:
>      I've learned much about postgres-xc recently. I am very 
> interested in its "architecture". I want do some experiments based on 
> it for my postgraduate paper.
>      By reading the documents and source code, I found it will be a 
> long time from postgres-xc stability and high performance. For 
> example, there's no any HA(High-Availability) solution such as standby 
> during the cluster running. In addition, many statements are not 
> supported in postgres-0.9.2-release.You will be greatly appreciated if 
> you could tell me something about postgres-xc recent and 
> future development schedule. According to that, I can make an 
> appropriate experiment plan. It’s a little bit urgent, thank you for 
> your reply. I will always keep eyes on our postgres-xc.
>                                                                           
> Yours truly,
>                                                                                
> Benny
>
Benny,

There was a talk we gave at PG con in Ottawa in May. In the presentation 
slides was some additional info about upcoming features. There is some 
audio here: http://www.pgcon.org/2010/schedule/events/226.en.html.

The amount of SQL coverage has expanded greatly compared to when we 
first announced the project, and the current development head adds even 
more, if you want to download that.

HA is currently being designed and discussed so you can expect to see 
that in the future.

If it fits in with your post-graduate work, please feel free to get 
involved and contribute to the project.

Regards,

Mason



>
> ------------------------------------------------------------------------------
> Start uncovering the many advantages of virtual appliances
> and start using them to simplify application deployment and
> accelerate your shift to cloud computing.
> http://p.sf.net/sfu/novell-sfdev2dev
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>    


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Dears:
     I've learned much about postgres-xc recently. I am very interested in its "architecture". I want do some experiments based on it for my postgraduate paper.
     By reading the documents and source code, I found it will be a long time from postgres-xc stability and high performance. For example, there's no any HA(High-Availability) solution such as standby during the cluster running. In addition, many statements are not supported in postgres-0.9.2-release.You will be greatly appreciated if you could tell me something about postgres-xc recent and future development schedule. According to that, I can make an appropriate experiment plan. It’s a little bit urgent, thank you for your reply. I will always keep eyes on our postgres-xc.
                                                                          Yours truly,
                                                                               Benny
 


      

Benny,

Thank you very much for the suggestion.    Yes, I understand the README is
not rich.   Instead, we have "architecture" document which you can download
separately, as well as other manual and guidelines.

Could you try these documents?   Especially, I made "architecture" document
as a separate one because XC internal is not simple to understand and I
thought usual document with graphics and tables will be more helpful.

Your advice is very valuable.   I'll be waiting for your further advice.

Thank you very much;
----------
Koichi Suzuki


2010/9/19 mei le <lem...@ya...>

> Dears :
>
>        This is Benny ,a postgraduate for DBMS from DUTchina. I have
> studied postgreSQL for a long time. And I have always kept close eye on
> postgreSQL cluster solutions such as PGCluster-2 and GridSQL etc. After
> postgres-xc released, I have learned much about it. It's a great cluster
> frame. I have been reading postgres-xc source code recently.
>
>
>
>          Reading PostgresSQL code, I always like to read README under
> relative directory, which can make me learn more knowledge on its relative
> functions. But when I scan postgres-xc code which was freshly added, I can't
> find any README to introduce the summary or relative knowledge. I think
> README is very important to an open source code. I don't know why
> postgres-xc ignores README files .Could we add the README files later?
>
>
>
>
> Truly yours,
>
>
>
> Benny
>
>
>
>
> ------------------------------------------------------------------------------
> Start uncovering the many advantages of virtual appliances
> and start using them to simplify application deployment and
> accelerate your shift to cloud computing.
> http://p.sf.net/sfu/novell-sfdev2dev
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>
>

Dears :
       This is Benny ,a postgraduate for DBMS from DUTchina. I have studied postgreSQL for a long time. And I have always kept close eye on postgreSQL cluster solutions such as PGCluster-2 and GridSQL etc. After postgres-xc released, I have learned much about it. It's a great cluster frame. I have been reading postgres-xc source code recently. 
 
         Reading PostgresSQL code, I always like to read README under relative directory, which can make me learn more knowledge on its relative functions. But when I scan postgres-xc code which was freshly added, I can't find any README to introduce the summary or relative knowledge. I think README is very important to an open source code. I don't know why postgres-xc ignores README files .Could we add the README files later?
 
                                                                                          Truly yours,
                                                                                           Benny



      

Thanks.

I'll collect other developer's opinion.  I see there's no harm anyway.

BTW, in terms of new rpm, we're planning to release the new version 
early in Octover with much more range of statement support, as well as 
synchronized DDL over multiple coordinators and other feature.

I think it's better to work with this version and I hope you can arrange 
your schedule more flexibly.  We're at the last stage of new version's 
release, which also comes with global installer/configurator.  I'll let 
you know the precise date.

Regards and thank you very much;
---
Koichi Suzuki

(2010年09月15日 15:47), Devrim GÜNDÜZ wrote:
> Hi,
>
> I remember that brought up this before:
>
> http://svn.pgrpms.org/browser/rpm/redhat/8.4/postgres-xc/F-13/postgres-xc-09-build.patch
>
> This patch is needed to build Postgres-XC under my boxes. If it does not
> break your environment, could you please apply it to git? Otherwise, I
> will keep it for RPMs.
>
> Regards,
>
>
>
>
> ------------------------------------------------------------------------------
> Start uncovering the many advantages of virtual appliances
> and start using them to simplify application deployment and
> accelerate your shift to cloud computing.
> http://p.sf.net/sfu/novell-sfdev2dev
>
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers

Hi,

I remember that brought up this before:

http://svn.pgrpms.org/browser/rpm/redhat/8.4/postgres-xc/F-13/postgres-xc-09-build.patch

This patch is needed to build Postgres-XC under my boxes. If it does not
break your environment, could you please apply it to git? Otherwise, I
will keep it for RPMs. 

Regards,

-- 
Devrim GÜNDÜZ
PostgreSQL Danışmanı/Consultant, Red Hat Certified Engineer
PostgreSQL RPM Repository: http://yum.pgrpms.org
Community: devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr
http://www.gunduz.org  Twitter: http://twitter.com/devrimgunduz

Andrei;

Thank you very much for detailed description of the spec and example test.

---
Koichi

(2010年06月23日 15:31), Andrei Martsinchyk wrote:
> Hi Sudo-san,
>
> Good description of CREATE AGGREGATE can be found in the Postgres docs:
>
> http://www.postgresql.org/docs/8.4/static/sql-createaggregate.html
>
> Postgres-XC added additional step (collection function) between
> transition and final calculation functions, and schema
>
> sfunc( internal-state, next-data-values ) --->  next-internal-state
> ffunc( internal-state ) --->  aggregate-value
>
> was changed to
>
> sfunc( node-internal-state, next-data-values ) --->  node-internal-state
> cfunc( coord-internal-state, node-internal-state ) --->  coord-internal-state
> ffunc( coord-internal-state ) --->  aggregate-value
>
> To reflect this change following parameters was added to CREATE AGGREGATE:
>
> CFUNC - collection function,
> CTYPE - collection data type,
> INITCOLLECT - (optional) initial collection value.
>
> It is hard to develop aggregate function from the scratch. For my
> testing I recreated existing functions under different name.
> Example:
>
> create aggregate mymin (int4) (sfunc=int4smaller, stype=int4,
> cfunc=int4smaller, ctype=int4);
>
> You can see the list of defined aggregate functions querying pg_aggregate table:
>
> select * pg_aggregate;
>
> You can also try and construct own aggregate function. You would need
> to choose functions for sfunc, cfunc and ffunc (component functions)
> from already defined in Postgres.
> You can see the list of defined functions querying pg_proc table:
>
> select * pg_proc;
>
> There are requirements for return types and parameter types of the
> component functions.
> cfunc:
> data type of the first parameter must be the same as return type. Data
> types of remaining parameters must be the same as data types of the
> aggregate  function
> sfunc:
> must have two parameters, data type of first parameter must be the
> same as return type of sfunc, data type of second parameter must be
> the same as return type of cfunc
> ffunc:
> must have one parameter, data type must be the same as return type of sfunc.
>
> And finally data types, you can see the list of defined data types
> querying pg_type table:
> select * pg_type;
>
> Feel free to ask if you have a question.
>

Hi Sudo-san,

Good description of CREATE AGGREGATE can be found in the Postgres docs:

http://www.postgresql.org/docs/8.4/static/sql-createaggregate.html

Postgres-XC added additional step (collection function) between
transition and final calculation functions, and schema

sfunc( internal-state, next-data-values ) ---> next-internal-state
ffunc( internal-state ) ---> aggregate-value

was changed to

sfunc( node-internal-state, next-data-values ) ---> node-internal-state
cfunc( coord-internal-state, node-internal-state ) ---> coord-internal-state
ffunc( coord-internal-state ) ---> aggregate-value

To reflect this change following parameters was added to CREATE AGGREGATE:

CFUNC - collection function,
CTYPE - collection data type,
INITCOLLECT - (optional) initial collection value.

It is hard to develop aggregate function from the scratch. For my
testing I recreated existing functions under different name.
Example:

create aggregate mymin (int4) (sfunc=int4smaller, stype=int4,
cfunc=int4smaller, ctype=int4);

You can see the list of defined aggregate functions querying pg_aggregate table:

select * pg_aggregate;

You can also try and construct own aggregate function. You would need
to choose functions for sfunc, cfunc and ffunc (component functions)
from already defined in Postgres.
You can see the list of defined functions querying pg_proc table:

select * pg_proc;

There are requirements for return types and parameter types of the
component functions.
cfunc:
data type of the first parameter must be the same as return type. Data
types of remaining parameters must be the same as data types of the
aggregate  function
sfunc:
must have two parameters, data type of first parameter must be the
same as return type of sfunc, data type of second parameter must be
the same as return type of cfunc
ffunc:
must have one parameter, data type must be the same as return type of sfunc.

And finally data types, you can see the list of defined data types
querying pg_type table:
select * pg_type;

Feel free to ask if you have a question.

-- 
Andrei Martsinchyk

EntepriseDB Corporation
The Enterprise Postgres Company

Website: www.enterprisedb.com
EnterpriseDB Blog: http://blogs.enterprisedb.com/
Follow us on Twitter: http://www.twitter.com/enterprisedb

This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Hi,

Perhaps you already saw this email in the postgres hackers mailing list...
My mistake... and my apologies...

Please see attached a patch that finishes the support of current values for
sequences.
The infrastructure was already in place, just a call to the correct API to
contact GTM was missing.

Regards,

-- 
Michael Paquier
NIPPON TELEGRAPH AND
TELEPHONE CORPORATION
NTT Open Source Software Center

If we do merge with 9.0, we still might want to merge in steps. to 8.4.4 
first, then 9.0 to more easily catch any problems and hopefully make the 
9.0 merge go more smoothly.

Unless we need any particular 9.0 feature, I am not in any hurry to 
merge until 9.0 has gone GA.  I like the idea of us working with a known 
stable base such that if we have a problem, we can be pretty sure it is 
in our code and not PostgreSQL. To be really paranoid, we could consider 
waiting until 9.0.1.

Thanks,

Mason


Koichi Suzuki wrote:
> Well, I think we should think more about merging to 9.0.
>
> ----------
> Koichi Suzuki
>
>
>
> 2010/5/26 Michael Paquier <mic...@gm...>:
>   
>> Hi,
>>
>> PostgreSQL 8.4.4 has been released on the 17th.
>> We didn't have any discussions about that yet, but it could be nice to do a
>> code merge.
>> Waiting for everybody's feedback.
>>
>> Regards,
>>
>> --
>> Michael Paquier
>> NIPPON TELEGRAPH AND
>> TELEPHONE CORPORATION
>> NTT Open Source Software Center
>>
>> ------------------------------------------------------------------------------
>>
>>
>> _______________________________________________
>> Postgres-xc-developers mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>>
>>
>>     
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>   


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Well, I think we should think more about merging to 9.0.

----------
Koichi Suzuki



2010/5/26 Michael Paquier <mic...@gm...>:
> Hi,
>
> PostgreSQL 8.4.4 has been released on the 17th.
> We didn't have any discussions about that yet, but it could be nice to do a
> code merge.
> Waiting for everybody's feedback.
>
> Regards,
>
> --
> Michael Paquier
> NIPPON TELEGRAPH AND
> TELEPHONE CORPORATION
> NTT Open Source Software Center
>
> ------------------------------------------------------------------------------
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>
>

Hi,

PostgreSQL 8.4.4 has been released on the 17th.
We didn't have any discussions about that yet, but it could be nice to do a
code merge.
Waiting for everybody's feedback.

Regards,

-- 
Michael Paquier
NIPPON TELEGRAPH AND
TELEPHONE CORPORATION
NTT Open Source Software Center

Ok Thanks,
I was a little bit lost with the grammar.
This is not yet documented in the reference manual and should be added.

-- 
Michael Paquier
NIPPON TELEGRAPH AND
TELEPHONE CORPORATION
NTT Open Source Software Center

Michael,

Try this:

execute direct on node 1 'select * from pg_proc where proname = ''foo''';

Command in EXECUTE DIRECT is not necessarily SELECT, but also any DDL
or DML statement.

2010/5/24 Michael Paquier <mic...@gm...>:
> Hi,
>
> I noticed a couple of days ago that it is actually not possible to execute
> directly to Datanodes a query
> with string elements. For example:
>
> 1) execute direct on node 1 'select * from pg_proc where proname = 'foo'';
> It doesn't work because the query is seen as a unique string for the parser.
> Parser on coordinator returns an error.
>
> 2)  execute direct on node 1 'select * from pg_proc where proname = foo';
> This one also doesn't work because foo has to be a string.
> Datanode returns an error.
>
> Here is my proposition to correct that:
> I don't know for the others, but basically I use execute direct only with
> select queries
> to see if data is correctly present on  data nodes, for debugging, etc.
> Why not replacing the string query 'Sconst' in EXECUTE DIRECT by a
> select_with_parens?
> This grammar is used by COPY (SELECT)
> With that, even if the use of execute direct is restricted to Select queries
> only,
> I feel that it becomes more useful.
>
> The new grammar would become:
> EXECUTE DIRECT ON
> { COORDINATOR | NODE list_of_nodes }
> ( SELECT query );
>
> Any opinions?
> Regards,
>
> --
> Michael Paquier
> NIPPON TELEGRAPH AND
> TELEPHONE CORPORATION
> NTT Open Source Software Center
>
> ------------------------------------------------------------------------------
>
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>
>



-- 
Andrei Martsinchyk

EntepriseDB Corporation
The Enterprise Postgres Company

Website: www.enterprisedb.com
EnterpriseDB Blog: http://blogs.enterprisedb.com/
Follow us on Twitter: http://www.twitter.com/enterprisedb

This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Hi,

I noticed a couple of days ago that it is actually not possible to execute
directly to Datanodes a query
with string elements. For example:

1) execute direct on node 1 'select * from pg_proc where proname = 'foo'';
It doesn't work because the query is seen as a unique string for the parser.
Parser on coordinator returns an error.

2)  execute direct on node 1 'select * from pg_proc where proname = foo';
This one also doesn't work because foo has to be a string.
Datanode returns an error.

Here is my proposition to correct that:
I don't know for the others, but basically I use execute direct only with
select queries
to see if data is correctly present on  data nodes, for debugging, etc.
Why not replacing the string query 'Sconst' in EXECUTE DIRECT by a
select_with_parens?
This grammar is used by COPY (SELECT)
With that, even if the use of execute direct is restricted to Select queries
only,
I feel that it becomes more useful.

The new grammar would become:
EXECUTE DIRECT ON
{ COORDINATOR | NODE list_of_nodes }
( SELECT query );

Any opinions?
Regards,

-- 
Michael Paquier
NIPPON TELEGRAPH AND
TELEPHONE CORPORATION
NTT Open Source Software Center

This is to announce Postgres-XC development team now provide
Postgres-XC 0.9.1 release at
https://sourceforge.net/projects/postgres-xc/

Information will be available in the project home page
http://postgres-xc.sourceforge.net/

Additional feature includes COPY FROM/TO, synchronous update of
replicated tables and simple aggregate functions.   The release
includes information how to run pgbench with Postgres-XC, as well as
its modified source code.

Postgres-XC replicate or distribute tables so that each statement can
run in a limited number of servers.   As a whole,  transactions can
run in PostgreSQL-XC nodes in parallel to achieve both read/write
scalability.     Global transaction management is provided so that any
update from any master is immediately visible from any other masters.
  You don't have to worry to distinguish read transaction from update
transaction.   You can issue both read/write transaction to any of the
master nodes.
----------
Postgres-XC development team,
Koichi Suzuki

Hi folks,

Just to let you know that I modified a little bit the configuration of the
git repository.
Now commit email looks cleaner. From now, code diff will appear for each
commit done.
I also deleted some annoying headers so as to keep the message as short as
possible with essential data only.

Regards,
-- 
Michael Paquier
NIPPON TELEGRAPH AND
TELEPHONE CORPORATION
NTT Open Source Software Center

Andrei updated the doc on aggregates, please see attached (thanks, 
Andrei!).

I forgot to mention when I committed his aggregate patch is that 
arrary() is not yet supported, but the other aggregate functions are 
(see doc for a full list).

Andrei also helpfully sent along COPY syntax:

COPY tablename [ ( column [, ...] ) ]
    FROM { 'filename' | STDIN }
    [ [ WITH ]
          [ OIDS ]
          [ DELIMITER [ AS ] 'delimiter' ]
          [ NULL [ AS ] 'null string' ]
          [ CSV [ HEADER ]
                [ QUOTE [ AS ] 'quote' ]
                [ ESCAPE [ AS ] 'escape' ]
                [ FORCE NOT NULL column [, ...] ]

COPY tablename [ ( column [, ...] ) ]
    TO { 'filename' | STDOUT }
    [ [ WITH ]
          [ OIDS ]
          [ DELIMITER [ AS ] 'delimiter' ]
          [ NULL [ AS ] 'null string' ]
          [ CSV [ HEADER ]
                [ QUOTE [ AS ] 'quote' ]
                [ ESCAPE [ AS ] 'escape' ]
                [ FORCE QUOTE column [, ...] ]


Comparing to Postgres, PGXC does not support BINARY.

Regards,

Mason


Koichi Suzuki wrote:
> Hi,
>
> I'm beginning to write a reference manual for the new feature.   I
> need the following info for this.
>
> 1. COPY TO/ COPY FROM statement definition,
> 2. pg_aggregate extension (columns, types and descriptions)
> 3. Any other changes to the catalog.
> 4. Supported aggregate function list.
>
> User function release is scheduled at the end of June so it will not
> be included yet.
>
> For the information of catalog, you may send me .h files.
>
> Also, I think we can add "aggregate function" algorithm to the
> release.   Would like to have a comment on this.
>
> Regards;
> ----------
> Koichi Suzuki
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>   


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.

Koichi Suzuki wrote:
> Hi Mason-san;
>
> Thanks a lot for the work.  I'd like to summarize what we can and what
> we cannot with PG-XC and pgadmin.
> ----------
> Koichi Suzuki
>   

With pgadmin, one can connect, browse the schema, tables, columns, view 
and edit data, etc. I just applied Andrei's aggregate patch, so SELECT 
COUNT(*) now works and that error no longer occurs as well.

Pgadmin is not "XC-aware", it just treats it as an 8.4 instance. That 
means you do not get an accurate picture of statistics and such on 
tables. We can later on have ANALYZE and VACUUM perhaps pull some of 
this info up from the data nodes and store the info in the 
coordinator(s) (or make these a view and pull data from the data 
nodes?). Long term we could also add distribution, replication, and 
perhaps even node information.

I did not spend too much time with pgadmin, but it is a smooth 
experience now. One thing that I noticed was it does not show defined 
sequences. Even if these are handled on GTM, we could update pg_catalog 
to view this information. I can create an issue for this on SF tracker.

Regards,

Mason


>
>
> 2010/5/13 Mason Sharp <mas...@en...>:
>   
>> Pavan Deolasee wrote:
>>     
>>> On Wed, May 12, 2010 at 3:19 AM, Mason Sharp
>>> <mas...@en... <mailto:mas...@en...>>
>>> wrote:
>>>
>>>     I started the task to recognize more "Postgres-XC safe" queries.
>>>
>>>     I initially focused on native pg_catalog based views like
>>>     pg_settings. This is because it used by pgadmin. With the attached
>>>     patch, pgadmin no longer complains about such queries because they
>>>     are now supported.
>>>
>>>     The PostgreSQL query rewriter appears to convert the view into a
>>>     subquery. I modified the XC planner to check for pg_catalog table
>>>     usage in the FROM clause (and subqueries thereof).
>>>
>>>
>>> FWIW here are couple of things that I noticed:
>>>
>>> +   foreach(item, rtable)
>>> +   {
>>> +       RangeTblEntry *rte = (RangeTblEntry *) lfirst(item);
>>> +
>>> +       if (rte->rtekind == RTE_RELATION)
>>> +       {
>>> +           if (rte->relid > FirstNormalObjectId)
>>> +               return false;
>>> +       } else if (rte->rtekind == RTE_SUBQUERY)
>>> +           return contains_only_pg_catalog (rte->subquery->rtable);
>>> +   }
>>>
>>> This code will break out of the loop if the subquery contains only
>>> pg_catalog tables. But other rtable entries may contain reference to
>>> non-system tables.
>>>
>>> The other thing I am worried about is whether its safe to just rely on
>>> FirstNormalObjectId to decide if a relation belongs to pg_catalog or
>>> not. A user can create a table in pg_catalog (though I think it
>>> requires changes to GUC) and that may get an OID larger than the
>>> FirstNormalObjectId or we may add a new default schema and create
>>> tables in the schema as part of initdb (we do that routinely in EDBAS
>>> for example). Should we just explicitly lookup the namespace of the
>>> relation instead ?
>>>
>>>       
>> Thanks, I made those changes and committed.
>>
>> Mason
>>
>>     
>>> Thanks,
>>> Pavan
>>>
>>> --
>>> Pavan Deolasee
>>> EnterpriseDB     http://www.enterprisedb.com
>>>       
>> --
>> Mason Sharp
>> EnterpriseDB Corporation
>> The Enterprise Postgres Company
>>
>>
>> This e-mail message (and any attachment) is intended for the use of
>> the individual or entity to whom it is addressed. This message
>> contains information from EnterpriseDB Corporation that may be
>> privileged, confidential, or exempt from disclosure under applicable
>> law. If you are not the intended recipient or authorized to receive
>> this for the intended recipient, any use, dissemination, distribution,
>> retention, archiving, or copying of this communication is strictly
>> prohibited. If you have received this e-mail in error, please notify
>> the sender immediately by reply e-mail and delete this message.
>>
>>
>> ------------------------------------------------------------------------------
>>
>> _______________________________________________
>> Postgres-xc-developers mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers
>>
>>     


-- 
Mason Sharp
EnterpriseDB Corporation
The Enterprise Postgres Company


This e-mail message (and any attachment) is intended for the use of
the individual or entity to whom it is addressed. This message
contains information from EnterpriseDB Corporation that may be
privileged, confidential, or exempt from disclosure under applicable
law. If you are not the intended recipient or authorized to receive
this for the intended recipient, any use, dissemination, distribution,
retention, archiving, or copying of this communication is strictly
prohibited. If you have received this e-mail in error, please notify
the sender immediately by reply e-mail and delete this message.