postgres-xc-general Mailing List for Postgres-XC (Page 4)

On Thu, Mar 14, 2013 at 7:20 PM, Arni Sumarlidason <
Arn...@md...> wrote:

>  *I ran this query,*
>
> select * from table t, table_lang l where t.id=l.id and l.date >=
> '2012-12-01' and l.date < '2013-01-01';****
>
> *with these results,*
>

Does this produce a crash without any data in the tables? If not, you need
to give the data as well. Please provide exact reproduction steps, which
when run without any modification, reproduce the crash. Also, you will need
to tell how the table is distributed, what XC configuration you are using
like how many datanodes/coordinators etc.


> **
>
> The connection to the server was lost. Attempting reset: Failed.****
>
> Time: 241861.411 ms****
>
> ** **
>
> Thank you for quick response,****
>
> Arni****
>
> ** **
>
> *From:* Ashutosh Bapat [mailto:ash...@en...]
> *Sent:* Thursday, March 14, 2013 11:17 AM
> *To:* Arni Sumarlidason
> *Cc:* pos...@li...
> *Subject:* Re: [Postgres-xc-general] Configuration Error - or?****
>
> ** **
>
> This should not happen, can you please provide reproduction steps.****
>
> On Thu, Mar 14, 2013 at 6:40 PM, Arni Sumarlidason <
> Arn...@md...> wrote:****
>
> Good Morning / Evening,****
>
>  ****
>
> I think I may have a configuration problem… I have data distributed
> relationally with date based partitions to 20 nodes. When I try to select
> it from a coordinator the query executes quickly and the data nodes
> promptly pump data to the coordinator. However I run out of memory(16bg)
> and the postgres process crashes. Is there some way to make the coordinator
> cache the data to its local table(local disk), or is there some way to get
> around this memory issue? ****
>
>  ****
>
> *Arni Sumarlidason | Software Engineer, Information Technology*****
>
> MDA | 820 West Diamond Ave | Gaithersburg, MD | USA****
>
> O: 240-833-8200 D: 240-833-8318 M: 256-393-2803****
>
> arn...@md...| http://www.mdaus.com<https://console.mxlogic.com/redir/?8VxNwQsL6zAQsCTNPb3a9EVd79I04GxHtenMTvANOoVcsCej76XCZEjrDmxfy6Hrc5v2vNmzaDE4endK3D7zqbbzaoVxdBwS-yr1vF6y0QJxqJ9Ao_gK7SbAvxYtfdbFEw4GMtAhrzIVlwq81GcqNd41sQglyidPYfDwedECSjrb38UsqenT3uvyc>
> ****
>
>  ****
>
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_mar<https://console.mxlogic.com/redir/?8VxNwQsL6zAQsCTNPb3a9EVd79I06JR7u1KktrPlBA2Q1l1d0llS9K_9zANOoVcsCedTdXgCTeJ2v4dmSoa-4_yJ6lfg8sKrs7ef6Qmn6kNP2rb1JZ4S2_id41Fr2Rqj8N-xsfIn8_3UWuqnjh09lwX8yT7pOH0Qg3koRyq82VEwH4ArDUvf0srjdICSm6hMUQsLK6VP9UQ-a4ueI>
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general<https://console.mxlogic.com/redir/?4sMUMqenzhOqejrUVBxB4QsCzASjDdqymovaAWtiHsruW01_wrxYGjFxYGjB1SKblFcz7W5M-NszYfzFVK_9zANOoVcsCedTdXgCTeJ2v4dmSoa-4_yJ6lfg8sKrs7ef6Qmn6kNP2rb1JZ4S2_id41Fr2Rqj8N-xsfIn8_3UWuqnjh09lwX8yT7pOH0Qg3koRyq82VEwH4ArDUvf0srpdICSm6hMUQsLK6QuvVWDEq>
> ****
>
>
>
>
> --
> Best Wishes,
> Ashutosh Bapat
> EntepriseDB Corporation
> The Enterprise Postgres Company****
>



-- 
Best Wishes,
Ashutosh Bapat
EntepriseDB Corporation
The Enterprise Postgres Company

I ran this query,
select * from table t, table_lang l where t.id=l.id and l.date >= '2012-12-01' and l.date < '2013-01-01';
with these results,
The connection to the server was lost. Attempting reset: Failed.
Time: 241861.411 ms

Thank you for quick response,
Arni

From: Ashutosh Bapat [mailto:ash...@en...]
Sent: Thursday, March 14, 2013 11:17 AM
To: Arni Sumarlidason
Cc: pos...@li...
Subject: Re: [Postgres-xc-general] Configuration Error - or?

This should not happen, can you please provide reproduction steps.
On Thu, Mar 14, 2013 at 6:40 PM, Arni Sumarlidason <Arn...@md...<mailto:Arn...@md...>> wrote:
Good Morning / Evening,

I think I may have a configuration problem... I have data distributed relationally with date based partitions to 20 nodes. When I try to select it from a coordinator the query executes quickly and the data nodes promptly pump data to the coordinator. However I run out of memory(16bg) and the postgres process crashes. Is there some way to make the coordinator cache the data to its local table(local disk), or is there some way to get around this memory issue?

Arni Sumarlidason | Software Engineer, Information Technology
MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
arn...@md...<mailto:arn...@md...>| http://www.mdaus.com<https://console.mxlogic.com/redir/?8VxNwQsL6zAQsCTNPb3a9EVd79I04GxHtenMTvANOoVcsCej76XCZEjrDmxfy6Hrc5v2vNmzaDE4endK3D7zqbbzaoVxdBwS-yr1vF6y0QJxqJ9Ao_gK7SbAvxYtfdbFEw4GMtAhrzIVlwq81GcqNd41sQglyidPYfDwedECSjrb38UsqenT3uvyc>


------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_mar<https://console.mxlogic.com/redir/?8VxNwQsL6zAQsCTNPb3a9EVd79I06JR7u1KktrPlBA2Q1l1d0llS9K_9zANOoVcsCedTdXgCTeJ2v4dmSoa-4_yJ6lfg8sKrs7ef6Qmn6kNP2rb1JZ4S2_id41Fr2Rqj8N-xsfIn8_3UWuqnjh09lwX8yT7pOH0Qg3koRyq82VEwH4ArDUvf0srjdICSm6hMUQsLK6VP9UQ-a4ueI>
_______________________________________________
Postgres-xc-general mailing list
Pos...@li...<mailto:Pos...@li...>
https://lists.sourceforge.net/lists/listinfo/postgres-xc-general<https://console.mxlogic.com/redir/?4sMUMqenzhOqejrUVBxB4QsCzASjDdqymovaAWtiHsruW01_wrxYGjFxYGjB1SKblFcz7W5M-NszYfzFVK_9zANOoVcsCedTdXgCTeJ2v4dmSoa-4_yJ6lfg8sKrs7ef6Qmn6kNP2rb1JZ4S2_id41Fr2Rqj8N-xsfIn8_3UWuqnjh09lwX8yT7pOH0Qg3koRyq82VEwH4ArDUvf0srpdICSm6hMUQsLK6QuvVWDEq>



--
Best Wishes,
Ashutosh Bapat
EntepriseDB Corporation
The Enterprise Postgres Company

This should not happen, can you please provide reproduction steps.

On Thu, Mar 14, 2013 at 6:40 PM, Arni Sumarlidason <
Arn...@md...> wrote:

>  Good Morning / Evening,****
>
> ** **
>
> I think I may have a configuration problem… I have data distributed
> relationally with date based partitions to 20 nodes. When I try to select
> it from a coordinator the query executes quickly and the data nodes
> promptly pump data to the coordinator. However I run out of memory(16bg)
> and the postgres process crashes. Is there some way to make the coordinator
> cache the data to its local table(local disk), or is there some way to get
> around this memory issue? ****
>
> ** **
>
> *Arni Sumarlidason | Software Engineer, Information Technology*****
>
> MDA | 820 West Diamond Ave | Gaithersburg, MD | USA****
>
> O: 240-833-8200 D: 240-833-8318 M: 256-393-2803****
>
> arn...@md...| http://www.mdaus.com  ****
>
> ** **
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_mar
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>
>


-- 
Best Wishes,
Ashutosh Bapat
EntepriseDB Corporation
The Enterprise Postgres Company

Good Morning / Evening,

I think I may have a configuration problem... I have data distributed relationally with date based partitions to 20 nodes. When I try to select it from a coordinator the query executes quickly and the data nodes promptly pump data to the coordinator. However I run out of memory(16bg) and the postgres process crashes. Is there some way to make the coordinator cache the data to its local table(local disk), or is there some way to get around this memory issue?

Arni Sumarlidason | Software Engineer, Information Technology
MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
arn...@md...<mailto:arn...@md...>| http://www.mdaus.com<http://www.mdaus.com/>

Good Morning everyone, I hope everyone had a good weekend,

I am attempting to cluster a table based on an index and I am receiving some weird messages. They don't seem to be fatal occurrences, but I am interested in what is causing them. Can anyone shed an light?

>From the terminal:
db_01=# cluster table_1207 using table_1207_pkey;
WARNING:  unexpected EOF on datanode connection
CLUSTER
Time: 5594.429 ms

>From pg_log on coord:
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  unexpected EOF on datanode connection
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  Unexpected data on connection, cleaning.
WARNING:  unexpected EOF on datanode connection

Arni Sumarlidason | Software Engineer, Information Technology
MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
arn...@md...<mailto:arn...@md...>| http://www.mdaus.com<http://www.mdaus.com/>

I see.   Installing XC resources in existing directory could be
harmful.   I think it is similar reason why pg_basebackup requires
empty directory to begin with.

pgxc_ctl has a means to clean up the target directory so it will help
to some extent.

Regards;
----------
Koichi Suzuki


2013/3/6 Theodotos Andreou <th...@ub...>:
> Guys just for the record I have found the solution on my original question.
>
> It seems that the GTM directory using the stock postgres-xc on Ubuntu was
> pre-initialised. After I created a new GTM directory and initialise it my
> self the error went away. This is from the logs:
>
>
> 1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Starting GTM server at
> (*:6666) -- control file /var/lib/postgres-xc/GTM-master/gtm.control
> LOCATION:  main,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:559
> 1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Restoring last GXID to
> 10000
>
> LOCATION:  GTM_RestoreTxnInfo,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_txn.c:2599
> 1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Started to run as
> GTM-Active.
> LOCATION:  main,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:631
>
> Thank you both for your time
>
>
>
> On 03/04/2013 06:34 PM, Koichi Suzuki wrote:
>>
>> Removing the target is just in case.   For example, pg_basebackup
>> requires the target directory empty.   It is better to copy only the
>> "needed" ones.   It is good thing to clear the resources to avoid
>> conflict with other materials.   If you'd like to eploy non-standard
>> binaries, say, contrib, you can install them locally.   It will be in
>> your $pgxcInstallDir directories, which automatically goes to all the
>> target servers with deploy command.
>>
>> Regards;
>> ----------
>> Koichi Suzuki
>>
>>
>> 2013/3/4 Nikhil Sontakke <ni...@st...>:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>>
>>>>
>>>> After deployment the above directories were apparently deleted:
>>>>
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> nodes  pgxcConf  pgxc_ctl_log
>>>>
>>>>
>>> Ok, it seems this node is part of your deploy targets as well!
>>>
>>>          "node13 node12 node06 node07 node08 node09"
>>>
>>> So that's why the deploy which first clears up existing directories
>>> removes them as well.
>>>
>>> Think of this node where you compiled the sources as the management
>>> node and use it to install binaries on other nodes. Obviously since
>>> the script is basically a bash script you should be able to modify it
>>> to check if this management node is also part of the deploy list as
>>> well (and submit it back as a patch) and not remove the directories in
>>> that case.
>>>
>>>
>>>> Do you by any chance have the answer to my original question? :) :
>>>>
>>> The message comes in from a node to register with the GTM. Are there
>>> any coordinator/datanode nodes already running when you start GTM? I
>>> would suggest that you get a clean cluster up first and then we can
>>> investigate further.
>>>
>>> Regards,
>>> Nikhils
>>>
>>>> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
>>>> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>>>>
>>>>
>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>> node not found in registered node(s).
>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>> startup message, but received �
>>>> LOCATION:  GTM_ThreadMain,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>>> state
>>>> LOCATION:  GTM_ThreadCleanup,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>
>>>>
>>>> If I can solve this I can proceed the installation step by step as per
>>>> this
>>>> guide:
>>>>
>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>
>>>> Does the above error means that I will have to set all services in the
>>>> cluster? Are there any other dependencies on the GTM?
>>>>
>>>>
>>>> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>>>>
>>>>> Hi Theodotos,
>>>>>
>>>>> What does
>>>>>
>>>>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>>>>
>>>>> Ensure that compile the pgxc sources with
>>>>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>>>>
>>>>> Regards,
>>>>> Nikhils
>>>>>
>>>>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>>>>> wrote:
>>>>>>
>>>>>> Hi Koichi, me again,
>>>>>>
>>>>>> I have build postgres-xc under $HOME/pgxc (In fact
>>>>>> /var/lib/postgres-xc/pgxc
>>>>>>
>>>>>> This is how the directory looks before running deploy:
>>>>>>
>>>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>>>>
>>>>>> When I run deploy I get:
>>>>>>
>>>>>> http://pastebin.com/nhgMrV8u
>>>>>>
>>>>>> tar fails to find bin, include etc again.
>>>>>>
>>>>>> Checking the contents of the directory I get:
>>>>>>
>>>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>>>> nodes  pgxcConf  pgxc_ctl_log
>>>>>>
>>>>>> Is this a bug in the script or is it me not following the correct
>>>>>> procedure? Please advice
>>>>>>
>>>>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>>>>
>>>>>>> OK I think I got it. This is not dowloading/building the binaries.
>>>>>>> You
>>>>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>>>>> Right?
>>>>>>>
>>>>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>>>>
>>>>>>>> Hi Koichi,
>>>>>>>>
>>>>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes
>>>>>>>> (node
>>>>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>>>>> postgres user from koichi to postgres-xc. The postgres-xc user
>>>>>>>> exists
>>>>>>>> on
>>>>>>>> all nodes and the control machine and there is passwordless
>>>>>>>> configuration on all the nodes from the control machine.
>>>>>>>>
>>>>>>>> This is my config as shown on "xcshow config"
>>>>>>>>
>>>>>>>> http://pastebin.com/Hiz5bEzw
>>>>>>>>
>>>>>>>>
>>>>>>>> When I run deploy all this is what I get:
>>>>>>>>
>>>>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>>>>> tar: bin: Cannot stat: No such file or directory
>>>>>>>> tar: include: Cannot stat: No such file or directory
>>>>>>>> tar: lib: Cannot stat: No such file or directory
>>>>>>>> tar: share: Cannot stat: No such file or directory
>>>>>>>> tar: Exiting with failure status due to previous errors
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>>
>>>>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as
>>>>>>>> root? I
>>>>>>>> tried as root but I get different errors
>>>>>>>>
>>>>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in
>>>>>>>> exe
>>>>>>>> right and it is owned by the postgres-xc user and group.
>>>>>>>>
>>>>>>>> Any ideas?
>>>>>>>>
>>>>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>>>>
>>>>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>>>>
>>>>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>>>>
>>>>>>>>>> Hello,
>>>>>>>>>>
>>>>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>>>>> ERROR messages in it.
>>>>>>>>>>
>>>>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be
>>>>>>>>>> found
>>>>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>>>>
>>>>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>>>>
>>>>>>>>>> I hope this helps.
>>>>>>>>>>
>>>>>>>>>> Regards;
>>>>>>>>>> ----------
>>>>>>>>>> Koichi Suzuki
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>>>>
>>>>>>>>>>> Hello to all
>>>>>>>>>>>
>>>>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>>>>> guide:
>>>>>>>>>>>
>>>>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>>>>
>>>>>>>>>>> I am at the first step, configuring the GTM master (page 39). I
>>>>>>>>>>> have
>>>>>>>>>>> this configuration:
>>>>>>>>>>>
>>>>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>>>>> listen_addresses = '*'
>>>>>>>>>>> port = 6666
>>>>>>>>>>> startup = ACT
>>>>>>>>>>>
>>>>>>>>>>> I run this command as postgres-xc:
>>>>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>>>>
>>>>>>>>>>> The server start apparently at port 6666:
>>>>>>>>>>>
>>>>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>>>>> 2408/gtm
>>>>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>>>>> 2408/gtm
>>>>>>>>>>>
>>>>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>>>>
>>>>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM
>>>>>>>>>>> standby
>>>>>>>>>>> node not found in registered node(s).
>>>>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting
>>>>>>>>>>> a
>>>>>>>>>>> startup message, but received �
>>>>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>>>>> thread
>>>>>>>>>>> state
>>>>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>>>>
>>>>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>>>>
>>>>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install
>>>>>>>>>>> postegres-xc
>>>>>>>>>>> from the repositories.
>>>>>>>>>>>
>>>>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>>>>
>>>>>>>>>>> Secondary question 1:
>>>>>>>>>>>
>>>>>>>>>>> In the guide it says:
>>>>>>>>>>> nodename = 'gtmName'
>>>>>>>>>>> for both master and standby. Does this imply that they should
>>>>>>>>>>> have
>>>>>>>>>>> the
>>>>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>>>>
>>>>>>>>>>> Secondary question 2:
>>>>>>>>>>>
>>>>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>>>>> switch
>>>>>>>>>>> the IP from master to slave using heartbeat or keepalived and
>>>>>>>>>>> avoid
>>>>>>>>>>> this
>>>>>>>>>>> step?
>>>>>>>>>>>
>>>>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>>>>> status
>>>>>>>>>>> is "newbie" :)
>>>>>>>>>>>
>>>>>>>>>>> Thanks
>>>>>>>>>>>
>>>>>>>>>>> Theo
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>>>>> _______________________________________________
>>>>>>>>>>> Postgres-xc-general mailing list
>>>>>>>>>>> Pos...@li...
>>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>>> _______________________________________________
>>>>>>>>> Postgres-xc-general mailing list
>>>>>>>>> Pos...@li...
>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>> _______________________________________________
>>>>>>>> Postgres-xc-general mailing list
>>>>>>>> Pos...@li...
>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Everyone hates slow websites. So do we.
>>>>>>> Make your web apps faster with AppDynamics
>>>>>>> Download AppDynamics Lite for free today:
>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>> _______________________________________________
>>>>>>> Postgres-xc-general mailing list
>>>>>>> Pos...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>>
>>>>>
>>>
>>>
>>> --
>>> StormDB - http://www.stormdb.com
>>> The Database Cloud
>>> Postgres-XC Support and Service
>
>

Guys just for the record I have found the solution on my original question.

It seems that the GTM directory using the stock postgres-xc on Ubuntu 
was pre-initialised. After I created a new GTM directory and initialise 
it my self the error went away. This is from the logs:


1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Starting GTM server 
at (*:6666) -- control file /var/lib/postgres-xc/GTM-master/gtm.control
LOCATION:  main, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:559
1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Restoring last GXID 
to 10000

LOCATION:  GTM_RestoreTxnInfo, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_txn.c:2599
1:139942913603392:2013-03-06 11:58:07.168 EET -LOG:  Started to run as 
GTM-Active.
LOCATION:  main, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:631

Thank you both for your time


On 03/04/2013 06:34 PM, Koichi Suzuki wrote:
> Removing the target is just in case.   For example, pg_basebackup
> requires the target directory empty.   It is better to copy only the
> "needed" ones.   It is good thing to clear the resources to avoid
> conflict with other materials.   If you'd like to eploy non-standard
> binaries, say, contrib, you can install them locally.   It will be in
> your $pgxcInstallDir directories, which automatically goes to all the
> target servers with deploy command.
>
> Regards;
> ----------
> Koichi Suzuki
>
>
> 2013/3/4 Nikhil Sontakke <ni...@st...>:
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>
>>>
>>> After deployment the above directories were apparently deleted:
>>>
>>>
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> nodes  pgxcConf  pgxc_ctl_log
>>>
>>>
>> Ok, it seems this node is part of your deploy targets as well!
>>
>>          "node13 node12 node06 node07 node08 node09"
>>
>> So that's why the deploy which first clears up existing directories
>> removes them as well.
>>
>> Think of this node where you compiled the sources as the management
>> node and use it to install binaries on other nodes. Obviously since
>> the script is basically a bash script you should be able to modify it
>> to check if this management node is also part of the deploy list as
>> well (and submit it back as a patch) and not remove the directories in
>> that case.
>>
>>
>>> Do you by any chance have the answer to my original question? :) :
>>>
>> The message comes in from a node to register with the GTM. Are there
>> any coordinator/datanode nodes already running when you start GTM? I
>> would suggest that you get a clean cluster up first and then we can
>> investigate further.
>>
>> Regards,
>> Nikhils
>>
>>> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
>>> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>>>
>>>
>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>> node not found in registered node(s).
>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>> startup message, but received �
>>> LOCATION:  GTM_ThreadMain,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>> state
>>> LOCATION:  GTM_ThreadCleanup,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>
>>>
>>> If I can solve this I can proceed the installation step by step as per this
>>> guide:
>>>
>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>
>>> Does the above error means that I will have to set all services in the
>>> cluster? Are there any other dependencies on the GTM?
>>>
>>>
>>> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>>> Hi Theodotos,
>>>>
>>>> What does
>>>>
>>>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>>>
>>>> Ensure that compile the pgxc sources with
>>>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>>>
>>>> Regards,
>>>> Nikhils
>>>>
>>>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>>>> wrote:
>>>>> Hi Koichi, me again,
>>>>>
>>>>> I have build postgres-xc under $HOME/pgxc (In fact
>>>>> /var/lib/postgres-xc/pgxc
>>>>>
>>>>> This is how the directory looks before running deploy:
>>>>>
>>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>>>
>>>>> When I run deploy I get:
>>>>>
>>>>> http://pastebin.com/nhgMrV8u
>>>>>
>>>>> tar fails to find bin, include etc again.
>>>>>
>>>>> Checking the contents of the directory I get:
>>>>>
>>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>>> nodes  pgxcConf  pgxc_ctl_log
>>>>>
>>>>> Is this a bug in the script or is it me not following the correct
>>>>> procedure? Please advice
>>>>>
>>>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>>> OK I think I got it. This is not dowloading/building the binaries. You
>>>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>>>> Right?
>>>>>>
>>>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>>> Hi Koichi,
>>>>>>>
>>>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists
>>>>>>> on
>>>>>>> all nodes and the control machine and there is passwordless
>>>>>>> configuration on all the nodes from the control machine.
>>>>>>>
>>>>>>> This is my config as shown on "xcshow config"
>>>>>>>
>>>>>>> http://pastebin.com/Hiz5bEzw
>>>>>>>
>>>>>>>
>>>>>>> When I run deploy all this is what I get:
>>>>>>>
>>>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>>>> tar: bin: Cannot stat: No such file or directory
>>>>>>> tar: include: Cannot stat: No such file or directory
>>>>>>> tar: lib: Cannot stat: No such file or directory
>>>>>>> tar: share: Cannot stat: No such file or directory
>>>>>>> tar: Exiting with failure status due to previous errors
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>>
>>>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>>>>> tried as root but I get different errors
>>>>>>>
>>>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>>>>> right and it is owned by the postgres-xc user and group.
>>>>>>>
>>>>>>> Any ideas?
>>>>>>>
>>>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>>>
>>>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>>> Hello,
>>>>>>>>>
>>>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>>>> ERROR messages in it.
>>>>>>>>>
>>>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>>>
>>>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>>>
>>>>>>>>> I hope this helps.
>>>>>>>>>
>>>>>>>>> Regards;
>>>>>>>>> ----------
>>>>>>>>> Koichi Suzuki
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>>> Hello to all
>>>>>>>>>>
>>>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>>>> guide:
>>>>>>>>>>
>>>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>>>
>>>>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>>>>> this configuration:
>>>>>>>>>>
>>>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>>>> listen_addresses = '*'
>>>>>>>>>> port = 6666
>>>>>>>>>> startup = ACT
>>>>>>>>>>
>>>>>>>>>> I run this command as postgres-xc:
>>>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>>>
>>>>>>>>>> The server start apparently at port 6666:
>>>>>>>>>>
>>>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>>>> 2408/gtm
>>>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>>>> 2408/gtm
>>>>>>>>>>
>>>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>>>
>>>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>>>>> node not found in registered node(s).
>>>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>>>
>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>>>>> startup message, but received �
>>>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>>>> thread
>>>>>>>>>> state
>>>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>>>
>>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>>>
>>>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>>>
>>>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>>>>> from the repositories.
>>>>>>>>>>
>>>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>>>
>>>>>>>>>> Secondary question 1:
>>>>>>>>>>
>>>>>>>>>> In the guide it says:
>>>>>>>>>> nodename = 'gtmName'
>>>>>>>>>> for both master and standby. Does this imply that they should have
>>>>>>>>>> the
>>>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>>>
>>>>>>>>>> Secondary question 2:
>>>>>>>>>>
>>>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>>>> switch
>>>>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid
>>>>>>>>>> this
>>>>>>>>>> step?
>>>>>>>>>>
>>>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>>>> status
>>>>>>>>>> is "newbie" :)
>>>>>>>>>>
>>>>>>>>>> Thanks
>>>>>>>>>>
>>>>>>>>>> Theo
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>>>> _______________________________________________
>>>>>>>>>> Postgres-xc-general mailing list
>>>>>>>>>> Pos...@li...
>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>>
>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>> _______________________________________________
>>>>>>>> Postgres-xc-general mailing list
>>>>>>>> Pos...@li...
>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Everyone hates slow websites. So do we.
>>>>>>> Make your web apps faster with AppDynamics
>>>>>>> Download AppDynamics Lite for free today:
>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>> _______________________________________________
>>>>>>> Postgres-xc-general mailing list
>>>>>>> Pos...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>
>>>>
>>
>>
>> --
>> StormDB - http://www.stormdb.com
>> The Database Cloud
>> Postgres-XC Support and Service

Removing the target is just in case.   For example, pg_basebackup
requires the target directory empty.   It is better to copy only the
"needed" ones.   It is good thing to clear the resources to avoid
conflict with other materials.   If you'd like to eploy non-standard
binaries, say, contrib, you can install them locally.   It will be in
your $pgxcInstallDir directories, which automatically goes to all the
target servers with deploy command.

Regards;
----------
Koichi Suzuki


2013/3/4 Nikhil Sontakke <ni...@st...>:
>>
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>
>>
>> After deployment the above directories were apparently deleted:
>>
>>
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> nodes  pgxcConf  pgxc_ctl_log
>>
>>
>
> Ok, it seems this node is part of your deploy targets as well!
>
>         "node13 node12 node06 node07 node08 node09"
>
> So that's why the deploy which first clears up existing directories
> removes them as well.
>
> Think of this node where you compiled the sources as the management
> node and use it to install binaries on other nodes. Obviously since
> the script is basically a bash script you should be able to modify it
> to check if this management node is also part of the deploy list as
> well (and submit it back as a patch) and not remove the directories in
> that case.
>
>
>> Do you by any chance have the answer to my original question? :) :
>>
>
> The message comes in from a node to register with the GTM. Are there
> any coordinator/datanode nodes already running when you start GTM? I
> would suggest that you get a clean cluster up first and then we can
> investigate further.
>
> Regards,
> Nikhils
>
>> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
>> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>>
>>
>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>> node not found in registered node(s).
>> LOCATION:  gtm_standby_connect_to_standby_int,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>> startup message, but received �
>> LOCATION:  GTM_ThreadMain,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>> state
>> LOCATION:  GTM_ThreadCleanup,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>
>>
>> If I can solve this I can proceed the installation step by step as per this
>> guide:
>>
>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>
>> Does the above error means that I will have to set all services in the
>> cluster? Are there any other dependencies on the GTM?
>>
>>
>> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>>
>>> Hi Theodotos,
>>>
>>> What does
>>>
>>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>>
>>> Ensure that compile the pgxc sources with
>>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>>
>>> Regards,
>>> Nikhils
>>>
>>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>>> wrote:
>>>>
>>>> Hi Koichi, me again,
>>>>
>>>> I have build postgres-xc under $HOME/pgxc (In fact
>>>> /var/lib/postgres-xc/pgxc
>>>>
>>>> This is how the directory looks before running deploy:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>>
>>>> When I run deploy I get:
>>>>
>>>> http://pastebin.com/nhgMrV8u
>>>>
>>>> tar fails to find bin, include etc again.
>>>>
>>>> Checking the contents of the directory I get:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> nodes  pgxcConf  pgxc_ctl_log
>>>>
>>>> Is this a bug in the script or is it me not following the correct
>>>> procedure? Please advice
>>>>
>>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>>
>>>>> OK I think I got it. This is not dowloading/building the binaries. You
>>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>>> Right?
>>>>>
>>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>>
>>>>>> Hi Koichi,
>>>>>>
>>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists
>>>>>> on
>>>>>> all nodes and the control machine and there is passwordless
>>>>>> configuration on all the nodes from the control machine.
>>>>>>
>>>>>> This is my config as shown on "xcshow config"
>>>>>>
>>>>>> http://pastebin.com/Hiz5bEzw
>>>>>>
>>>>>>
>>>>>> When I run deploy all this is what I get:
>>>>>>
>>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>>> tar: bin: Cannot stat: No such file or directory
>>>>>> tar: include: Cannot stat: No such file or directory
>>>>>> tar: lib: Cannot stat: No such file or directory
>>>>>> tar: share: Cannot stat: No such file or directory
>>>>>> tar: Exiting with failure status due to previous errors
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>
>>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>>>> tried as root but I get different errors
>>>>>>
>>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>>>> right and it is owned by the postgres-xc user and group.
>>>>>>
>>>>>> Any ideas?
>>>>>>
>>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>>
>>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>>
>>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>>
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>>> ERROR messages in it.
>>>>>>>>
>>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>>
>>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>>
>>>>>>>> I hope this helps.
>>>>>>>>
>>>>>>>> Regards;
>>>>>>>> ----------
>>>>>>>> Koichi Suzuki
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>>
>>>>>>>>> Hello to all
>>>>>>>>>
>>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>>> guide:
>>>>>>>>>
>>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>>
>>>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>>>> this configuration:
>>>>>>>>>
>>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>>> listen_addresses = '*'
>>>>>>>>> port = 6666
>>>>>>>>> startup = ACT
>>>>>>>>>
>>>>>>>>> I run this command as postgres-xc:
>>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>>
>>>>>>>>> The server start apparently at port 6666:
>>>>>>>>>
>>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>>> 2408/gtm
>>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>>> 2408/gtm
>>>>>>>>>
>>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>>
>>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>>>> node not found in registered node(s).
>>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>>
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>>>> startup message, but received �
>>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>>> thread
>>>>>>>>> state
>>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>>
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>>
>>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>>
>>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>>>> from the repositories.
>>>>>>>>>
>>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>>
>>>>>>>>> Secondary question 1:
>>>>>>>>>
>>>>>>>>> In the guide it says:
>>>>>>>>> nodename = 'gtmName'
>>>>>>>>> for both master and standby. Does this imply that they should have
>>>>>>>>> the
>>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>>
>>>>>>>>> Secondary question 2:
>>>>>>>>>
>>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>>> switch
>>>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid
>>>>>>>>> this
>>>>>>>>> step?
>>>>>>>>>
>>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>>> status
>>>>>>>>> is "newbie" :)
>>>>>>>>>
>>>>>>>>> Thanks
>>>>>>>>>
>>>>>>>>> Theo
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>>> _______________________________________________
>>>>>>>>> Postgres-xc-general mailing list
>>>>>>>>> Pos...@li...
>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Everyone hates slow websites. So do we.
>>>>>>> Make your web apps faster with AppDynamics
>>>>>>> Download AppDynamics Lite for free today:
>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>> _______________________________________________
>>>>>>> Postgres-xc-general mailing list
>>>>>>> Pos...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>
>>>
>>>
>>
>
>
>
> --
> StormDB - http://www.stormdb.com
> The Database Cloud
> Postgres-XC Support and Service

Actually when I tried the manual method I only setup just the GTM. No 
other nodes were present. I will try setup the whole cluster according 
to that HA guide and let you know.

Regarding the pgxc_ctl script I think I have followed all the guidelines 
you guys suggested. Passwordless ssh using the postgre-xc user, keeping 
the deploy server out of the cluster, setting the correct directory but 
I still get fails. I will revert to the manual method and if that does 
not work as expected we can revisit pgxc_ctl.

Thanks for your time guys.

On 03/04/2013 08:55 AM, Nikhil Sontakke wrote:
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>
>>
>> After deployment the above directories were apparently deleted:
>>
>>
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> nodes  pgxcConf  pgxc_ctl_log
>>
>>
> Ok, it seems this node is part of your deploy targets as well!
>
>          "node13 node12 node06 node07 node08 node09"
>
> So that's why the deploy which first clears up existing directories
> removes them as well.
>
> Think of this node where you compiled the sources as the management
> node and use it to install binaries on other nodes. Obviously since
> the script is basically a bash script you should be able to modify it
> to check if this management node is also part of the deploy list as
> well (and submit it back as a patch) and not remove the directories in
> that case.
>
>
>> Do you by any chance have the answer to my original question? :) :
>>
> The message comes in from a node to register with the GTM. Are there
> any coordinator/datanode nodes already running when you start GTM? I
> would suggest that you get a clean cluster up first and then we can
> investigate further.
>
> Regards,
> Nikhils
>
>> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
>> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>>
>>
>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>> node not found in registered node(s).
>> LOCATION:  gtm_standby_connect_to_standby_int,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>> startup message, but received �
>> LOCATION:  GTM_ThreadMain,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>> state
>> LOCATION:  GTM_ThreadCleanup,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>
>>
>> If I can solve this I can proceed the installation step by step as per this
>> guide:
>>
>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>
>> Does the above error means that I will have to set all services in the
>> cluster? Are there any other dependencies on the GTM?
>>
>>
>> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>> Hi Theodotos,
>>>
>>> What does
>>>
>>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>>
>>> Ensure that compile the pgxc sources with
>>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>>
>>> Regards,
>>> Nikhils
>>>
>>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>>> wrote:
>>>> Hi Koichi, me again,
>>>>
>>>> I have build postgres-xc under $HOME/pgxc (In fact
>>>> /var/lib/postgres-xc/pgxc
>>>>
>>>> This is how the directory looks before running deploy:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>>
>>>> When I run deploy I get:
>>>>
>>>> http://pastebin.com/nhgMrV8u
>>>>
>>>> tar fails to find bin, include etc again.
>>>>
>>>> Checking the contents of the directory I get:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>>> nodes  pgxcConf  pgxc_ctl_log
>>>>
>>>> Is this a bug in the script or is it me not following the correct
>>>> procedure? Please advice
>>>>
>>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>> OK I think I got it. This is not dowloading/building the binaries. You
>>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>>> Right?
>>>>>
>>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>> Hi Koichi,
>>>>>>
>>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists
>>>>>> on
>>>>>> all nodes and the control machine and there is passwordless
>>>>>> configuration on all the nodes from the control machine.
>>>>>>
>>>>>> This is my config as shown on "xcshow config"
>>>>>>
>>>>>> http://pastebin.com/Hiz5bEzw
>>>>>>
>>>>>>
>>>>>> When I run deploy all this is what I get:
>>>>>>
>>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>>> tar: bin: Cannot stat: No such file or directory
>>>>>> tar: include: Cannot stat: No such file or directory
>>>>>> tar: lib: Cannot stat: No such file or directory
>>>>>> tar: share: Cannot stat: No such file or directory
>>>>>> tar: Exiting with failure status due to previous errors
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>>
>>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>>>> tried as root but I get different errors
>>>>>>
>>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>>>> right and it is owned by the postgres-xc user and group.
>>>>>>
>>>>>> Any ideas?
>>>>>>
>>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>>
>>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>>> ERROR messages in it.
>>>>>>>>
>>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>>
>>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>>
>>>>>>>> I hope this helps.
>>>>>>>>
>>>>>>>> Regards;
>>>>>>>> ----------
>>>>>>>> Koichi Suzuki
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>> Hello to all
>>>>>>>>>
>>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>>> guide:
>>>>>>>>>
>>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>>
>>>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>>>> this configuration:
>>>>>>>>>
>>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>>> listen_addresses = '*'
>>>>>>>>> port = 6666
>>>>>>>>> startup = ACT
>>>>>>>>>
>>>>>>>>> I run this command as postgres-xc:
>>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>>
>>>>>>>>> The server start apparently at port 6666:
>>>>>>>>>
>>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>>> 2408/gtm
>>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>>> 2408/gtm
>>>>>>>>>
>>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>>
>>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>>>> node not found in registered node(s).
>>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>>
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>>>> startup message, but received �
>>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>>> thread
>>>>>>>>> state
>>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>>
>>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>>
>>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>>
>>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>>>> from the repositories.
>>>>>>>>>
>>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>>
>>>>>>>>> Secondary question 1:
>>>>>>>>>
>>>>>>>>> In the guide it says:
>>>>>>>>> nodename = 'gtmName'
>>>>>>>>> for both master and standby. Does this imply that they should have
>>>>>>>>> the
>>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>>
>>>>>>>>> Secondary question 2:
>>>>>>>>>
>>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>>> switch
>>>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid
>>>>>>>>> this
>>>>>>>>> step?
>>>>>>>>>
>>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>>> status
>>>>>>>>> is "newbie" :)
>>>>>>>>>
>>>>>>>>> Thanks
>>>>>>>>>
>>>>>>>>> Theo
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>>> _______________________________________________
>>>>>>>>> Postgres-xc-general mailing list
>>>>>>>>> Pos...@li...
>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Everyone hates slow websites. So do we.
>>>>>>> Make your web apps faster with AppDynamics
>>>>>>> Download AppDynamics Lite for free today:
>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>> _______________________________________________
>>>>>>> Postgres-xc-general mailing list
>>>>>>> Pos...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>
>>>
>
>

>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>
>
> After deployment the above directories were apparently deleted:
>
>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> nodes  pgxcConf  pgxc_ctl_log
>
>

Ok, it seems this node is part of your deploy targets as well!

        "node13 node12 node06 node07 node08 node09"

So that's why the deploy which first clears up existing directories
removes them as well.

Think of this node where you compiled the sources as the management
node and use it to install binaries on other nodes. Obviously since
the script is basically a bash script you should be able to modify it
to check if this management node is also part of the deploy list as
well (and submit it back as a patch) and not remove the directories in
that case.


> Do you by any chance have the answer to my original question? :) :
>

The message comes in from a node to register with the GTM. Are there
any coordinator/datanode nodes already running when you start GTM? I
would suggest that you get a clean cluster up first and then we can
investigate further.

Regards,
Nikhils

> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>
>
> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
> node not found in registered node(s).
> LOCATION:  gtm_standby_connect_to_standby_int,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
> startup message, but received �
> LOCATION:  GTM_ThreadMain,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
> state
> LOCATION:  GTM_ThreadCleanup,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>
>
> If I can solve this I can proceed the installation step by step as per this
> guide:
>
> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>
> Does the above error means that I will have to set all services in the
> cluster? Are there any other dependencies on the GTM?
>
>
> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>
>> Hi Theodotos,
>>
>> What does
>>
>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>
>> Ensure that compile the pgxc sources with
>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>
>> Regards,
>> Nikhils
>>
>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>> wrote:
>>>
>>> Hi Koichi, me again,
>>>
>>> I have build postgres-xc under $HOME/pgxc (In fact
>>> /var/lib/postgres-xc/pgxc
>>>
>>> This is how the directory looks before running deploy:
>>>
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>
>>> When I run deploy I get:
>>>
>>> http://pastebin.com/nhgMrV8u
>>>
>>> tar fails to find bin, include etc again.
>>>
>>> Checking the contents of the directory I get:
>>>
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> nodes  pgxcConf  pgxc_ctl_log
>>>
>>> Is this a bug in the script or is it me not following the correct
>>> procedure? Please advice
>>>
>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>
>>>> OK I think I got it. This is not dowloading/building the binaries. You
>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>> Right?
>>>>
>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>
>>>>> Hi Koichi,
>>>>>
>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists
>>>>> on
>>>>> all nodes and the control machine and there is passwordless
>>>>> configuration on all the nodes from the control machine.
>>>>>
>>>>> This is my config as shown on "xcshow config"
>>>>>
>>>>> http://pastebin.com/Hiz5bEzw
>>>>>
>>>>>
>>>>> When I run deploy all this is what I get:
>>>>>
>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>> tar: bin: Cannot stat: No such file or directory
>>>>> tar: include: Cannot stat: No such file or directory
>>>>> tar: lib: Cannot stat: No such file or directory
>>>>> tar: share: Cannot stat: No such file or directory
>>>>> tar: Exiting with failure status due to previous errors
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>
>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>>> tried as root but I get different errors
>>>>>
>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>>> right and it is owned by the postgres-xc user and group.
>>>>>
>>>>> Any ideas?
>>>>>
>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>
>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>
>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>> ERROR messages in it.
>>>>>>>
>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>
>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>
>>>>>>> I hope this helps.
>>>>>>>
>>>>>>> Regards;
>>>>>>> ----------
>>>>>>> Koichi Suzuki
>>>>>>>
>>>>>>>
>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>
>>>>>>>> Hello to all
>>>>>>>>
>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>> guide:
>>>>>>>>
>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>
>>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>>> this configuration:
>>>>>>>>
>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>> listen_addresses = '*'
>>>>>>>> port = 6666
>>>>>>>> startup = ACT
>>>>>>>>
>>>>>>>> I run this command as postgres-xc:
>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>
>>>>>>>> The server start apparently at port 6666:
>>>>>>>>
>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>> 2408/gtm
>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>> 2408/gtm
>>>>>>>>
>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>
>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>>> node not found in registered node(s).
>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>>> startup message, but received �
>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>> thread
>>>>>>>> state
>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>
>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>
>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>>> from the repositories.
>>>>>>>>
>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>
>>>>>>>> Secondary question 1:
>>>>>>>>
>>>>>>>> In the guide it says:
>>>>>>>> nodename = 'gtmName'
>>>>>>>> for both master and standby. Does this imply that they should have
>>>>>>>> the
>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>
>>>>>>>> Secondary question 2:
>>>>>>>>
>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>> switch
>>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid
>>>>>>>> this
>>>>>>>> step?
>>>>>>>>
>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>> status
>>>>>>>> is "newbie" :)
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>> Theo
>>>>>>>>
>>>>>>>>
>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>> _______________________________________________
>>>>>>>> Postgres-xc-general mailing list
>>>>>>>> Pos...@li...
>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>
>>
>>
>



-- 
StormDB - http://www.stormdb.com
The Database Cloud
Postgres-XC Support and Service

Andreou;

Here's a couple of tips to use pgxc_ctl and it will also be useful in
configuring XC.  Because XC spreads across more than one server
machine and because it is not good that root owns the database
resource, pgxc_ctl assumes not to use "root".   Also, it is highly
recommended to set up key-based authentication between servers.
Pgxc_ctl depends upon scp and ssh so key-based authentication
eliminates the need to type your password again and again.   Anyway,
you cannot/should not use "root" across the servers by scp and root.
This is quite bothering to setup such key-based authentication for all
the path.   I'm using the attached script (set_autologin) for this,
which works at least on Ubuntu and Centos and saves some of the time.

Pgxc_ctl requires two user names, $pgxcUser and $pgxcOwner.   The
former is Linux user name who owns all the Postgres-XC resouces.   The
latter is Postgres-XC database super user name you're using.   In my
sample configuration (you can get using prepare command, as found in
manual.txt), both are "koichi".   You can choose whatever you like but
please understand the previous paragraph.

Before you "deploy" postgreSQL binaries, you must have them in the
local server you're running pgxc_ctl.   This is typically the
directory name you specify with --prefix option in ./configure
command.   pgxc_ctl assumes all the binaries, libraries, included
files and all the templatefiles in the same directory.   If you don't
specify --prefix option in ./configure command, install directory will
be /usr/local/pgsql as specified by vanilla PostgreSQL binary build.

You should specify this local installation directory as the value of
$pgxcInstallDir.    This value is used both in your local server
running pgxc_ctl and other servers you're installing postgres-XC
binaries.

Please note that all the materials will be installes in
$pgxcInstallDir/{bin|lib|include|share}.   You should specify these
directories as a part of your $PATH and $LD_LIBRARY_PATH environment
in all the servers running Postgres-XC nodes
(gtm/gtm_proxy/coordinator/datanode).

Then you can specify the rest of your XC configuration to pgxc_ctl
configuration file.   Yes, you can edit pgxc_ctl file directly but it
can easily mess things up.   Instead, please use "prepare config"
command to create your configuration file.    It is safer to edit
this.

Default full path of the configuration file is
$pgxcIndtallDir/pgxcConf.   Please edit plgxc_ctl directory to change
this.

I do hope these help.   Please do feel free to write to me if you have
any other issues.

Best Regards;
----------
Koichi Suzuki


2013/3/4 Theodotos Andreou <th...@ub...>:
> After I build the system it was showing this:
>
>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>
>
> After deployment the above directories were apparently deleted:
>
>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> nodes  pgxcConf  pgxc_ctl_log
>
>
> postgres-xc was compiled with "--prefix=/var/lib/postgres-xc/pgxc" and "make
> install" finished successfully.
>
> Do you by any chance have the answer to my original question? :) :
>
> Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D
> /var/lib/postgres-xc/GTM start ) I get this in the logs:
>
>
> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
> node not found in registered node(s).
> LOCATION:  gtm_standby_connect_to_standby_int,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
> startup message, but received �
> LOCATION:  GTM_ThreadMain,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
> state
> LOCATION:  GTM_ThreadCleanup,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>
>
> If I can solve this I can proceed the installation step by step as per this
> guide:
>
> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>
> Does the above error means that I will have to set all services in the
> cluster? Are there any other dependencies on the GTM?
>
>
> On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
>>
>> Hi Theodotos,
>>
>> What does
>>
>> "ls /var/lib/postgres-xc/pgxc" show on your system?
>>
>> Ensure that compile the pgxc sources with
>> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>>
>> Regards,
>> Nikhils
>>
>> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...>
>> wrote:
>>>
>>> Hi Koichi, me again,
>>>
>>> I have build postgres-xc under $HOME/pgxc (In fact
>>> /var/lib/postgres-xc/pgxc
>>>
>>> This is how the directory looks before running deploy:
>>>
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>>
>>> When I run deploy I get:
>>>
>>> http://pastebin.com/nhgMrV8u
>>>
>>> tar fails to find bin, include etc again.
>>>
>>> Checking the contents of the directory I get:
>>>
>>> postgres-xc@pgxc-ctl:~$ ls pgxc
>>> nodes  pgxcConf  pgxc_ctl_log
>>>
>>> Is this a bug in the script or is it me not following the correct
>>> procedure? Please advice
>>>
>>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>>>
>>>> OK I think I got it. This is not dowloading/building the binaries. You
>>>> have to do that yourself! And then deploy will send it to the nodes.
>>>> Right?
>>>>
>>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>>>
>>>>> Hi Koichi,
>>>>>
>>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists
>>>>> on
>>>>> all nodes and the control machine and there is passwordless
>>>>> configuration on all the nodes from the control machine.
>>>>>
>>>>> This is my config as shown on "xcshow config"
>>>>>
>>>>> http://pastebin.com/Hiz5bEzw
>>>>>
>>>>>
>>>>> When I run deploy all this is what I get:
>>>>>
>>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>>> tar: bin: Cannot stat: No such file or directory
>>>>> tar: include: Cannot stat: No such file or directory
>>>>> tar: lib: Cannot stat: No such file or directory
>>>>> tar: share: Cannot stat: No such file or directory
>>>>> tar: Exiting with failure status due to previous errors
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>>
>>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>>> tried as root but I get different errors
>>>>>
>>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>>> right and it is owned by the postgres-xc user and group.
>>>>>
>>>>> Any ideas?
>>>>>
>>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>>>
>>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>>
>>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>>> ERROR messages in it.
>>>>>>>
>>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>>
>>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>>
>>>>>>> I hope this helps.
>>>>>>>
>>>>>>> Regards;
>>>>>>> ----------
>>>>>>> Koichi Suzuki
>>>>>>>
>>>>>>>
>>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>>>
>>>>>>>> Hello to all
>>>>>>>>
>>>>>>>> I am trying to setup a HA postgres-xc cluster according to this
>>>>>>>> guide:
>>>>>>>>
>>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>>
>>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>>> this configuration:
>>>>>>>>
>>>>>>>> nodename = 'mygtmnode01'
>>>>>>>> listen_addresses = '*'
>>>>>>>> port = 6666
>>>>>>>> startup = ACT
>>>>>>>>
>>>>>>>> I run this command as postgres-xc:
>>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>>
>>>>>>>> The server start apparently at port 6666:
>>>>>>>>
>>>>>>>> # netstat -lnptu | grep gtm
>>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN
>>>>>>>> 2408/gtm
>>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN
>>>>>>>> 2408/gtm
>>>>>>>>
>>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>>
>>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>>> node not found in registered node(s).
>>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>>>
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>>> startup message, but received �
>>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up
>>>>>>>> thread
>>>>>>>> state
>>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>>>
>>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>>
>>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>>
>>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>>> from the repositories.
>>>>>>>>
>>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>>
>>>>>>>> Secondary question 1:
>>>>>>>>
>>>>>>>> In the guide it says:
>>>>>>>> nodename = 'gtmName'
>>>>>>>> for both master and standby. Does this imply that they should have
>>>>>>>> the
>>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>>
>>>>>>>> Secondary question 2:
>>>>>>>>
>>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just
>>>>>>>> switch
>>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid
>>>>>>>> this
>>>>>>>> step?
>>>>>>>>
>>>>>>>> You would probably have figure it out already that my postgres-xc
>>>>>>>> status
>>>>>>>> is "newbie" :)
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>> Theo
>>>>>>>>
>>>>>>>>
>>>>>>>> ------------------------------------------------------------------------------
>>>>>>>> Everyone hates slow websites. So do we.
>>>>>>>> Make your web apps faster with AppDynamics
>>>>>>>> Download AppDynamics Lite for free today:
>>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>>> _______________________________________________
>>>>>>>> Postgres-xc-general mailing list
>>>>>>>> Pos...@li...
>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>
>>
>>
>

After I build the system it was showing this:

postgres-xc@pgxc-ctl:~$ ls pgxc
bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share


After deployment the above directories were apparently deleted:

postgres-xc@pgxc-ctl:~$ ls pgxc
nodes  pgxcConf  pgxc_ctl_log


postgres-xc was compiled with "--prefix=/var/lib/postgres-xc/pgxc" and 
"make install" finished successfully.

Do you by any chance have the answer to my original question? :) :

Why when I use gtm from Ubuntu repos (gtm_ctl -Z gtm -D 
/var/lib/postgres-xc/GTM start ) I get this in the logs:

1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
node not found in registered node(s).
LOCATION:  gtm_standby_connect_to_standby_int,
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
startup message, but received �
LOCATION:  GTM_ThreadMain,
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
state
LOCATION:  GTM_ThreadCleanup,
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265


If I can solve this I can proceed the installation step by step as per 
this guide:

http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf

Does the above error means that I will have to set all services in the 
cluster? Are there any other dependencies on the GTM?

On 03/03/2013 05:21 PM, Nikhil Sontakke wrote:
> Hi Theodotos,
>
> What does
>
> "ls /var/lib/postgres-xc/pgxc" show on your system?
>
> Ensure that compile the pgxc sources with
> --prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"
>
> Regards,
> Nikhils
>
> On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...> wrote:
>> Hi Koichi, me again,
>>
>> I have build postgres-xc under $HOME/pgxc (In fact /var/lib/postgres-xc/pgxc
>>
>> This is how the directory looks before running deploy:
>>
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>>
>> When I run deploy I get:
>>
>> http://pastebin.com/nhgMrV8u
>>
>> tar fails to find bin, include etc again.
>>
>> Checking the contents of the directory I get:
>>
>> postgres-xc@pgxc-ctl:~$ ls pgxc
>> nodes  pgxcConf  pgxc_ctl_log
>>
>> Is this a bug in the script or is it me not following the correct
>> procedure? Please advice
>>
>> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>>> OK I think I got it. This is not dowloading/building the binaries. You
>>> have to do that yourself! And then deploy will send it to the nodes. Right?
>>>
>>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>>> Hi Koichi,
>>>>
>>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>>> postgres user from koichi to postgres-xc. The postgres-xc user exists on
>>>> all nodes and the control machine and there is passwordless
>>>> configuration on all the nodes from the control machine.
>>>>
>>>> This is my config as shown on "xcshow config"
>>>>
>>>> http://pastebin.com/Hiz5bEzw
>>>>
>>>>
>>>> When I run deploy all this is what I get:
>>>>
>>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>>> tar: bin: Cannot stat: No such file or directory
>>>> tar: include: Cannot stat: No such file or directory
>>>> tar: lib: Cannot stat: No such file or directory
>>>> tar: share: Cannot stat: No such file or directory
>>>> tar: Exiting with failure status due to previous errors
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>>
>>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>>> tried as root but I get different errors
>>>>
>>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>>> right and it is owned by the postgres-xc user and group.
>>>>
>>>> Any ideas?
>>>>
>>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>>
>>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>>> Hello,
>>>>>>
>>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>>> ERROR messages in it.
>>>>>>
>>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>>
>>>>>> This is bash script so you can find what to do for XC cluster
>>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>>
>>>>>> I hope this helps.
>>>>>>
>>>>>> Regards;
>>>>>> ----------
>>>>>> Koichi Suzuki
>>>>>>
>>>>>>
>>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>>> Hello to all
>>>>>>>
>>>>>>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>>>>>>
>>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>>
>>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>>> this configuration:
>>>>>>>
>>>>>>> nodename = 'mygtmnode01'
>>>>>>> listen_addresses = '*'
>>>>>>> port = 6666
>>>>>>> startup = ACT
>>>>>>>
>>>>>>> I run this command as postgres-xc:
>>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>>
>>>>>>> The server start apparently at port 6666:
>>>>>>>
>>>>>>> # netstat -lnptu | grep gtm
>>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>>>>>>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>>>>>>
>>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>>
>>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>>> node not found in registered node(s).
>>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>>> startup message, but received �
>>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>>>>>> state
>>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>>
>>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>>
>>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>>> from the repositories.
>>>>>>>
>>>>>>> I haven't set up the GTM standy yet.
>>>>>>>
>>>>>>> Secondary question 1:
>>>>>>>
>>>>>>> In the guide it says:
>>>>>>> nodename = 'gtmName'
>>>>>>> for both master and standby. Does this imply that they should have the
>>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>>
>>>>>>> Secondary question 2:
>>>>>>>
>>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>>> reconfigure the proxy to the new master (ex standby). Can we just switch
>>>>>>> the IP from master to slave using heartbeat or keepalived and avoid this
>>>>>>> step?
>>>>>>>
>>>>>>> You would probably have figure it out already that my postgres-xc status
>>>>>>> is "newbie" :)
>>>>>>>
>>>>>>> Thanks
>>>>>>>
>>>>>>> Theo
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Everyone hates slow websites. So do we.
>>>>>>> Make your web apps faster with AppDynamics
>>>>>>> Download AppDynamics Lite for free today:
>>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>>> _______________________________________________
>>>>>>> Postgres-xc-general mailing list
>>>>>>> Pos...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>
>

Hi Theodotos,

What does

"ls /var/lib/postgres-xc/pgxc" show on your system?

Ensure that compile the pgxc sources with
--prefix=/var/lib/postgres-xc/pgxc and don't forget to "make install"

Regards,
Nikhils

On Sun, Mar 3, 2013 at 7:18 PM, Theodotos Andreou <th...@ub...> wrote:
> Hi Koichi, me again,
>
> I have build postgres-xc under $HOME/pgxc (In fact /var/lib/postgres-xc/pgxc
>
> This is how the directory looks before running deploy:
>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share
>
> When I run deploy I get:
>
> http://pastebin.com/nhgMrV8u
>
> tar fails to find bin, include etc again.
>
> Checking the contents of the directory I get:
>
> postgres-xc@pgxc-ctl:~$ ls pgxc
> nodes  pgxcConf  pgxc_ctl_log
>
> Is this a bug in the script or is it me not following the correct
> procedure? Please advice
>
> On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
>> OK I think I got it. This is not dowloading/building the binaries. You
>> have to do that yourself! And then deploy will send it to the nodes. Right?
>>
>> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>>> Hi Koichi,
>>>
>>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>>> separate machine to run pgxc_ctl on. The only thing I change was the
>>> postgres user from koichi to postgres-xc. The postgres-xc user exists on
>>> all nodes and the control machine and there is passwordless
>>> configuration on all the nodes from the control machine.
>>>
>>> This is my config as shown on "xcshow config"
>>>
>>> http://pastebin.com/Hiz5bEzw
>>>
>>>
>>> When I run deploy all this is what I get:
>>>
>>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>>> tar: bin: Cannot stat: No such file or directory
>>> tar: include: Cannot stat: No such file or directory
>>> tar: lib: Cannot stat: No such file or directory
>>> tar: share: Cannot stat: No such file or directory
>>> tar: Exiting with failure status due to previous errors
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>> wk.tgz 100%   45     0.0KB/s   00:00
>>>
>>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>>> tried as root but I get different errors
>>>
>>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>>> right and it is owned by the postgres-xc user and group.
>>>
>>> Any ideas?
>>>
>>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>>
>>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>>> Hello,
>>>>>
>>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>>> ERROR messages in it.
>>>>>
>>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>>
>>>>> This is bash script so you can find what to do for XC cluster
>>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>>
>>>>> I hope this helps.
>>>>>
>>>>> Regards;
>>>>> ----------
>>>>> Koichi Suzuki
>>>>>
>>>>>
>>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>>> Hello to all
>>>>>>
>>>>>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>>>>>
>>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>>
>>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>>> this configuration:
>>>>>>
>>>>>> nodename = 'mygtmnode01'
>>>>>> listen_addresses = '*'
>>>>>> port = 6666
>>>>>> startup = ACT
>>>>>>
>>>>>> I run this command as postgres-xc:
>>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>>
>>>>>> The server start apparently at port 6666:
>>>>>>
>>>>>> # netstat -lnptu | grep gtm
>>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>>>>>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>>>>>
>>>>>> But checking the logs I get repeatedly the following messages:
>>>>>>
>>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>>> node not found in registered node(s).
>>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>>> startup message, but received �
>>>>>> LOCATION:  GTM_ThreadMain,
>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>>>>> state
>>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>>
>>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>>
>>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>>> from the repositories.
>>>>>>
>>>>>> I haven't set up the GTM standy yet.
>>>>>>
>>>>>> Secondary question 1:
>>>>>>
>>>>>> In the guide it says:
>>>>>> nodename = 'gtmName'
>>>>>> for both master and standby. Does this imply that they should have the
>>>>>> same node name? Does it have to be the same as the hostname?
>>>>>>
>>>>>> Secondary question 2:
>>>>>>
>>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>>> reconfigure the proxy to the new master (ex standby). Can we just switch
>>>>>> the IP from master to slave using heartbeat or keepalived and avoid this
>>>>>> step?
>>>>>>
>>>>>> You would probably have figure it out already that my postgres-xc status
>>>>>> is "newbie" :)
>>>>>>
>>>>>> Thanks
>>>>>>
>>>>>> Theo
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> Everyone hates slow websites. So do we.
>>>>>> Make your web apps faster with AppDynamics
>>>>>> Download AppDynamics Lite for free today:
>>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>>> _______________________________________________
>>>>>> Postgres-xc-general mailing list
>>>>>> Pos...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general



-- 
StormDB - http://www.stormdb.com
The Database Cloud
Postgres-XC Support and Service

Hi Koichi, me again,

I have build postgres-xc under $HOME/pgxc (In fact /var/lib/postgres-xc/pgxc

This is how the directory looks before running deploy:

postgres-xc@pgxc-ctl:~$ ls pgxc
bin  include  lib  nodes  pgxcConf  pgxc_ctl_log  share

When I run deploy I get:

http://pastebin.com/nhgMrV8u

tar fails to find bin, include etc again.

Checking the contents of the directory I get:

postgres-xc@pgxc-ctl:~$ ls pgxc
nodes  pgxcConf  pgxc_ctl_log

Is this a bug in the script or is it me not following the correct 
procedure? Please advice

On 03/03/2013 03:14 PM, Theodotos Andreou wrote:
> OK I think I got it. This is not dowloading/building the binaries. You
> have to do that yourself! And then deploy will send it to the nodes. Right?
>
> On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
>> Hi Koichi,
>>
>> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
>> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
>> separate machine to run pgxc_ctl on. The only thing I change was the
>> postgres user from koichi to postgres-xc. The postgres-xc user exists on
>> all nodes and the control machine and there is passwordless
>> configuration on all the nodes from the control machine.
>>
>> This is my config as shown on "xcshow config"
>>
>> http://pastebin.com/Hiz5bEzw
>>
>>
>> When I run deploy all this is what I get:
>>
>> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
>> tar: bin: Cannot stat: No such file or directory
>> tar: include: Cannot stat: No such file or directory
>> tar: lib: Cannot stat: No such file or directory
>> tar: share: Cannot stat: No such file or directory
>> tar: Exiting with failure status due to previous errors
>> wk.tgz 100%   45     0.0KB/s   00:00
>> wk.tgz 100%   45     0.0KB/s   00:00
>> wk.tgz 100%   45     0.0KB/s   00:00
>> wk.tgz 100%   45     0.0KB/s   00:00
>> wk.tgz 100%   45     0.0KB/s   00:00
>> wk.tgz 100%   45     0.0KB/s   00:00
>>
>> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
>> tried as root but I get different errors
>>
>> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
>> right and it is owned by the postgres-xc user and group.
>>
>> Any ideas?
>>
>> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>>> Thanks for the tip. I' ll try that and be back with more feedback
>>>
>>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>>> Hello,
>>>>
>>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>>> ERROR messages in it.
>>>>
>>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>>
>>>> This is bash script so you can find what to do for XC cluster
>>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>>
>>>> I hope this helps.
>>>>
>>>> Regards;
>>>> ----------
>>>> Koichi Suzuki
>>>>
>>>>
>>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>>> Hello to all
>>>>>
>>>>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>>>>
>>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>>
>>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>>> this configuration:
>>>>>
>>>>> nodename = 'mygtmnode01'
>>>>> listen_addresses = '*'
>>>>> port = 6666
>>>>> startup = ACT
>>>>>
>>>>> I run this command as postgres-xc:
>>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>>
>>>>> The server start apparently at port 6666:
>>>>>
>>>>> # netstat -lnptu | grep gtm
>>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>>>>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>>>>
>>>>> But checking the logs I get repeatedly the following messages:
>>>>>
>>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>>> node not found in registered node(s).
>>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>>> startup message, but received �
>>>>> LOCATION:  GTM_ThreadMain,
>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>>>> state
>>>>> LOCATION:  GTM_ThreadCleanup,
>>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>>
>>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>>
>>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>>> from the repositories.
>>>>>
>>>>> I haven't set up the GTM standy yet.
>>>>>
>>>>> Secondary question 1:
>>>>>
>>>>> In the guide it says:
>>>>> nodename = 'gtmName'
>>>>> for both master and standby. Does this imply that they should have the
>>>>> same node name? Does it have to be the same as the hostname?
>>>>>
>>>>> Secondary question 2:
>>>>>
>>>>> In the GTM proxy procedure when the master fails it suggests to
>>>>> reconfigure the proxy to the new master (ex standby). Can we just switch
>>>>> the IP from master to slave using heartbeat or keepalived and avoid this
>>>>> step?
>>>>>
>>>>> You would probably have figure it out already that my postgres-xc status
>>>>> is "newbie" :)
>>>>>
>>>>> Thanks
>>>>>
>>>>> Theo
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Everyone hates slow websites. So do we.
>>>>> Make your web apps faster with AppDynamics
>>>>> Download AppDynamics Lite for free today:
>>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>>> _______________________________________________
>>>>> Postgres-xc-general mailing list
>>>>> Pos...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

OK I think I got it. This is not dowloading/building the binaries. You 
have to do that yourself! And then deploy will send it to the nodes. Right?

On 03/03/2013 08:34 AM, Theodotos Andreou wrote:
> Hi Koichi,
>
> So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node
> 12 - 13) as the default configuration of pgxc_ctl. I also setup a
> separate machine to run pgxc_ctl on. The only thing I change was the
> postgres user from koichi to postgres-xc. The postgres-xc user exists on
> all nodes and the control machine and there is passwordless
> configuration on all the nodes from the control machine.
>
> This is my config as shown on "xcshow config"
>
> http://pastebin.com/Hiz5bEzw
>
>
> When I run deploy all this is what I get:
>
> postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
> tar: bin: Cannot stat: No such file or directory
> tar: include: Cannot stat: No such file or directory
> tar: lib: Cannot stat: No such file or directory
> tar: share: Cannot stat: No such file or directory
> tar: Exiting with failure status due to previous errors
> wk.tgz 100%   45     0.0KB/s   00:00
> wk.tgz 100%   45     0.0KB/s   00:00
> wk.tgz 100%   45     0.0KB/s   00:00
> wk.tgz 100%   45     0.0KB/s   00:00
> wk.tgz 100%   45     0.0KB/s   00:00
> wk.tgz 100%   45     0.0KB/s   00:00
>
> Am I supposed to run "deploy all" as the postgres-xc user or as root? I
> tried as root but I get different errors
>
> The ~/bin directory does in fact exist with the pgxc_ctl binary in exe
> right and it is owned by the postgres-xc user and group.
>
> Any ideas?
>
> On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
>> Thanks for the tip. I' ll try that and be back with more feedback
>>
>> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>>> Hello,
>>>
>>> THanks a lot for the mail.   A log as I see, there*s no proble in
>>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>>> ERROR messages in it.
>>>
>>> Could you try pgxc_ctl to configure your cluster?   It will be found
>>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>>
>>> This is bash script so you can find what to do for XC cluster
>>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>>
>>> I hope this helps.
>>>
>>> Regards;
>>> ----------
>>> Koichi Suzuki
>>>
>>>
>>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>>> Hello to all
>>>>
>>>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>>>
>>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>>
>>>> I am at the first step, configuring the GTM master (page 39). I have
>>>> this configuration:
>>>>
>>>> nodename = 'mygtmnode01'
>>>> listen_addresses = '*'
>>>> port = 6666
>>>> startup = ACT
>>>>
>>>> I run this command as postgres-xc:
>>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>>
>>>> The server start apparently at port 6666:
>>>>
>>>> # netstat -lnptu | grep gtm
>>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>>>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>>>
>>>> But checking the logs I get repeatedly the following messages:
>>>>
>>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>>> node not found in registered node(s).
>>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>>> startup message, but received �
>>>> LOCATION:  GTM_ThreadMain,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>>> state
>>>> LOCATION:  GTM_ThreadCleanup,
>>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>>
>>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>>
>>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>>> from the repositories.
>>>>
>>>> I haven't set up the GTM standy yet.
>>>>
>>>> Secondary question 1:
>>>>
>>>> In the guide it says:
>>>> nodename = 'gtmName'
>>>> for both master and standby. Does this imply that they should have the
>>>> same node name? Does it have to be the same as the hostname?
>>>>
>>>> Secondary question 2:
>>>>
>>>> In the GTM proxy procedure when the master fails it suggests to
>>>> reconfigure the proxy to the new master (ex standby). Can we just switch
>>>> the IP from master to slave using heartbeat or keepalived and avoid this
>>>> step?
>>>>
>>>> You would probably have figure it out already that my postgres-xc status
>>>> is "newbie" :)
>>>>
>>>> Thanks
>>>>
>>>> Theo
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics
>>>> Download AppDynamics Lite for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

Hi Koichi,

So I have setup 4 coord/datanodes (node06 - 09) and two gtm nodes (node 
12 - 13) as the default configuration of pgxc_ctl. I also setup a 
separate machine to run pgxc_ctl on. The only thing I change was the 
postgres user from koichi to postgres-xc. The postgres-xc user exists on 
all nodes and the control machine and there is passwordless 
configuration on all the nodes from the control machine.

This is my config as shown on "xcshow config"

http://pastebin.com/Hiz5bEzw


When I run deploy all this is what I get:

postgres-xc@pgxc-ctl:~$ pgxc_ctl deploy all
tar: bin: Cannot stat: No such file or directory
tar: include: Cannot stat: No such file or directory
tar: lib: Cannot stat: No such file or directory
tar: share: Cannot stat: No such file or directory
tar: Exiting with failure status due to previous errors
wk.tgz 100%   45     0.0KB/s   00:00
wk.tgz 100%   45     0.0KB/s   00:00
wk.tgz 100%   45     0.0KB/s   00:00
wk.tgz 100%   45     0.0KB/s   00:00
wk.tgz 100%   45     0.0KB/s   00:00
wk.tgz 100%   45     0.0KB/s   00:00

Am I supposed to run "deploy all" as the postgres-xc user or as root? I 
tried as root but I get different errors

The ~/bin directory does in fact exist with the pgxc_ctl binary in exe 
right and it is owned by the postgres-xc user and group.

Any ideas?

On 03/02/2013 07:03 PM, Theodotos Andreou wrote:
> Thanks for the tip. I' ll try that and be back with more feedback
>
> On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
>> Hello,
>>
>> THanks a lot for the mail.   A log as I see, there*s no proble in
>> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
>> ERROR messages in it.
>>
>> Could you try pgxc_ctl to configure your cluster?   It will be found
>> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>>
>> This is bash script so you can find what to do for XC cluster
>> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>>
>> I hope this helps.
>>
>> Regards;
>> ----------
>> Koichi Suzuki
>>
>>
>> 2013/3/2 Theodotos Andreou <th...@ub...>:
>>> Hello to all
>>>
>>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>>
>>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>>
>>> I am at the first step, configuring the GTM master (page 39). I have
>>> this configuration:
>>>
>>> nodename = 'mygtmnode01'
>>> listen_addresses = '*'
>>> port = 6666
>>> startup = ACT
>>>
>>> I run this command as postgres-xc:
>>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>>
>>> The server start apparently at port 6666:
>>>
>>> # netstat -lnptu | grep gtm
>>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>>
>>> But checking the logs I get repeatedly the following messages:
>>>
>>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>>> node not found in registered node(s).
>>> LOCATION:  gtm_standby_connect_to_standby_int,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>>> startup message, but received �
>>> LOCATION:  GTM_ThreadMain,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>>> state
>>> LOCATION:  GTM_ThreadCleanup,
>>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>>
>>> That FATAL error above is scaring me. Am I doing something wrong?
>>>
>>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>>> from the repositories.
>>>
>>> I haven't set up the GTM standy yet.
>>>
>>> Secondary question 1:
>>>
>>> In the guide it says:
>>> nodename = 'gtmName'
>>> for both master and standby. Does this imply that they should have the
>>> same node name? Does it have to be the same as the hostname?
>>>
>>> Secondary question 2:
>>>
>>> In the GTM proxy procedure when the master fails it suggests to
>>> reconfigure the proxy to the new master (ex standby). Can we just switch
>>> the IP from master to slave using heartbeat or keepalived and avoid this
>>> step?
>>>
>>> You would probably have figure it out already that my postgres-xc status
>>> is "newbie" :)
>>>
>>> Thanks
>>>
>>> Theo
>>>
>>> ------------------------------------------------------------------------------
>>> Everyone hates slow websites. So do we.
>>> Make your web apps faster with AppDynamics
>>> Download AppDynamics Lite for free today:
>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>> _______________________________________________
>>> Postgres-xc-general mailing list
>>> Pos...@li...
>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

Thanks for the tip. I' ll try that and be back with more feedback

On 03/02/2013 03:56 AM, Koichi Suzuki wrote:
> Hello,
>
> THanks a lot for the mail.   A log as I see, there*s no proble in
> gtm.conf file.   I looked into my gtm.log and did see any FATAL or
> ERROR messages in it.
>
> Could you try pgxc_ctl to configure your cluster?   It will be found
> in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl
>
> This is bash script so you can find what to do for XC cluster
> operation.   Attached is my gtm.conf generated by pgxc_ctl.
>
> I hope this helps.
>
> Regards;
> ----------
> Koichi Suzuki
>
>
> 2013/3/2 Theodotos Andreou <th...@ub...>:
>> Hello to all
>>
>> I am trying to setup a HA postgres-xc cluster according to this guide:
>>
>> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>>
>> I am at the first step, configuring the GTM master (page 39). I have
>> this configuration:
>>
>> nodename = 'mygtmnode01'
>> listen_addresses = '*'
>> port = 6666
>> startup = ACT
>>
>> I run this command as postgres-xc:
>> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>>
>> The server start apparently at port 6666:
>>
>> # netstat -lnptu | grep gtm
>> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
>> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>>
>> But checking the logs I get repeatedly the following messages:
>>
>> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
>> node not found in registered node(s).
>> LOCATION:  gtm_standby_connect_to_standby_int,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
>> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
>> startup message, but received �
>> LOCATION:  GTM_ThreadMain,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
>> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
>> state
>> LOCATION:  GTM_ThreadCleanup,
>> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>>
>> That FATAL error above is scaring me. Am I doing something wrong?
>>
>> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
>> from the repositories.
>>
>> I haven't set up the GTM standy yet.
>>
>> Secondary question 1:
>>
>> In the guide it says:
>> nodename = 'gtmName'
>> for both master and standby. Does this imply that they should have the
>> same node name? Does it have to be the same as the hostname?
>>
>> Secondary question 2:
>>
>> In the GTM proxy procedure when the master fails it suggests to
>> reconfigure the proxy to the new master (ex standby). Can we just switch
>> the IP from master to slave using heartbeat or keepalived and avoid this
>> step?
>>
>> You would probably have figure it out already that my postgres-xc status
>> is "newbie" :)
>>
>> Thanks
>>
>> Theo
>>
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

>
> We are running metrics on the system now, we loaded some data recently, and
> continuing to do so now. If Hint bits are the case, what steps can I take to
> relieve some of this IO?
>

This is internal IO which will settle down if you are not loading into
the same table again immediately. A lumpsum count(1) call (after a
recent data load) will touch all pages immediately, so try to avoid
that and amortize the IO over multiple specific select calls.

HTH,
Nikhils


>
>
> From: Mason Sharp [mailto:ma...@st...]
> Sent: Friday, March 01, 2013 2:29 PM
> To: Arni Sumarlidason
> Cc: pos...@li...; Postgres-XC Developers
> (pos...@li...)
> Subject: Re: [Postgres-xc-general] IO PGXC
>
>
>
>
> Sent from my IPhone
>
>
> On Mar 1, 2013, at 4:13 PM, Arni Sumarlidason <Arn...@md...>
> wrote:
>
> Users,
>
>
>
> I have 20 nodes sitting on a disk arrays, with multiple LUNs. when I issue
> queries – `select count(1) from table` for example, I am experiencing heavy
> writes and heavy reads. I expected the reads but not the writes and it has
> really thrown a wrench in the caching. My first assumption would be the log
> files, do you have any other ideas what could be causing all these writes
> with a select?
>
>
>
> It could be vacuum. Also, did you just load data? Hint bits get updated,
> dirtying pages and causing them to be written.
>
>
>
>
>
> Arni Sumarlidason | Software Engineer, Information Technology
>
> MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
>
> O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
>
> arn...@md...| http://www.mdaus.com
>
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
>
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general



-- 
StormDB - http://www.stormdb.com
The Database Cloud
Postgres-XC Support and Service

Hello,

THanks a lot for the mail.   A log as I see, there*s no proble in
gtm.conf file.   I looked into my gtm.log and did see any FATAL or
ERROR messages in it.

Could you try pgxc_ctl to configure your cluster?   It will be found
in https://github.com/koichi-szk/PGXC-Tools/tree/master/pgxc_ctl

This is bash script so you can find what to do for XC cluster
operation.   Attached is my gtm.conf generated by pgxc_ctl.

I hope this helps.

Regards;
----------
Koichi Suzuki


2013/3/2 Theodotos Andreou <th...@ub...>:
> Hello to all
>
> I am trying to setup a HA postgres-xc cluster according to this guide:
>
> http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf
>
> I am at the first step, configuring the GTM master (page 39). I have
> this configuration:
>
> nodename = 'mygtmnode01'
> listen_addresses = '*'
> port = 6666
> startup = ACT
>
> I run this command as postgres-xc:
> $ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start
>
> The server start apparently at port 6666:
>
> # netstat -lnptu | grep gtm
> tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
> tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm
>
> But checking the logs I get repeatedly the following messages:
>
> 1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby
> node not found in registered node(s).
> LOCATION:  gtm_standby_connect_to_standby_int,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
> 1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a
> startup message, but received �
> LOCATION:  GTM_ThreadMain,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
> 1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread
> state
> LOCATION:  GTM_ThreadCleanup,
> /build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265
>
> That FATAL error above is scaring me. Am I doing something wrong?
>
> I am running a 64 bit ubuntu (13.04) and I have install postegres-xc
> from the repositories.
>
> I haven't set up the GTM standy yet.
>
> Secondary question 1:
>
> In the guide it says:
> nodename = 'gtmName'
> for both master and standby. Does this imply that they should have the
> same node name? Does it have to be the same as the hostname?
>
> Secondary question 2:
>
> In the GTM proxy procedure when the master fails it suggests to
> reconfigure the proxy to the new master (ex standby). Can we just switch
> the IP from master to slave using heartbeat or keepalived and avoid this
> step?
>
> You would probably have figure it out already that my postgres-xc status
> is "newbie" :)
>
> Thanks
>
> Theo
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

Hello to all

I am trying to setup a HA postgres-xc cluster according to this guide:

http://wiki.postgresql.org/images/4/44/Pgxc_HA_20121024.pdf

I am at the first step, configuring the GTM master (page 39). I have 
this configuration:

nodename = 'mygtmnode01'
listen_addresses = '*'
port = 6666
startup = ACT

I run this command as postgres-xc:
$ gtm_ctl -Z gtm -D /var/lib/postgres-xc/GTM start

The server start apparently at port 6666:

# netstat -lnptu | grep gtm
tcp        0      0 0.0.0.0:6666            0.0.0.0:* LISTEN      2408/gtm
tcp6       0      0 :::6666                 :::* LISTEN      2408/gtm

But checking the logs I get repeatedly the following messages:

1:139871852988224:2013-03-01 19:28:36.528 EET -LOG:  Any GTM standby 
node not found in registered node(s).
LOCATION:  gtm_standby_connect_to_standby_int, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_standby.c:378
1:139871844607744:2013-03-01 19:28:36.528 EET -FATAL:  Expecting a 
startup message, but received �
LOCATION:  GTM_ThreadMain, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/main.c:985
1:139871844607744:2013-03-01 19:28:36.529 EET -LOG:  Cleaning up thread 
state
LOCATION:  GTM_ThreadCleanup, 
/build/buildd/postgres-xc-1.0.2/build/../src/gtm/main/gtm_thread.c:265

That FATAL error above is scaring me. Am I doing something wrong?

I am running a 64 bit ubuntu (13.04) and I have install postegres-xc 
from the repositories.

I haven't set up the GTM standy yet.

Secondary question 1:

In the guide it says:
nodename = 'gtmName'
for both master and standby. Does this imply that they should have the 
same node name? Does it have to be the same as the hostname?

Secondary question 2:

In the GTM proxy procedure when the master fails it suggests to 
reconfigure the proxy to the new master (ex standby). Can we just switch 
the IP from master to slave using heartbeat or keepalived and avoid this 
step?

You would probably have figure it out already that my postgres-xc status 
is "newbie" :)

Thanks

Theo

Thank you for your fast responses,

We are running metrics on the system now, we loaded some data recently, and continuing to do so now. If Hint bits are the case, what steps can I take to relieve some of this IO?

From: Mason Sharp [mailto:ma...@st...]
Sent: Friday, March 01, 2013 2:29 PM
To: Arni Sumarlidason
Cc: pos...@li...; Postgres-XC Developers (pos...@li...)
Subject: Re: [Postgres-xc-general] IO PGXC


Sent from my IPhone

On Mar 1, 2013, at 4:13 PM, Arni Sumarlidason <Arn...@md...<mailto:Arn...@md...>> wrote:
Users,

I have 20 nodes sitting on a disk arrays, with multiple LUNs. when I issue queries – `select count(1) from table` for example, I am experiencing heavy writes and heavy reads. I expected the reads but not the writes and it has really thrown a wrench in the caching. My first assumption would be the log files, do you have any other ideas what could be causing all these writes with a select?

It could be vacuum. Also, did you just load data? Hint bits get updated, dirtying pages and causing them to be written.



Arni Sumarlidason | Software Engineer, Information Technology
MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
arn...@md...<mailto:arn...@md...>| http://www.mdaus.com<https://console.mxlogic.com/redir/?4sMUMqenzhOqejrUVBxB4QsCzAS02lgRKDbUrLOrb0VVdOXMUTsSHJaM1l9RyGwgVsSYOYMrhpsjhsshpvdIICMnWhEwdbomHip6fQbxZyV7Uv7jPiWq80Qc6y0b181x8TfM-u0USyrpdIIczxNEVvsdD8rO>

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_feb<https://console.mxlogic.com/redir/?8VxNwQsL6zAQsCTNPb3a9EVd79I06JR7u1KktrPlBA2Q1l1d0ltYwrLOrb0VVdOXMUTsSHJaM1l9RyGwgVsSYOYMrhpsjhsshpvdIICMnWhEwdbomHip6fQbxZyV7Uv7jPiWq80Qc6y0b181x8TfM-u0USCrpdIIczxNEVvsdBa9a>
_______________________________________________
Postgres-xc-general mailing list
Pos...@li...<mailto:Pos...@li...>
https://lists.sourceforge.net/lists/listinfo/postgres-xc-general<https://console.mxlogic.com/redir/?2eosod7bNEVd79JYsOMOyqejhOr9PCJhbcfBiteFlKdLt00_MdM-l9QM-l9OwXn5GQChzZ2UvoKh-7NQYTvASm1POrBTxNKVJnqlw2GjH5l0xOVJVBVwSyOUCyUUyO-rppdwLQzh0qmMJmAOcvEn3X5OfM-eDCBQQg1Eod40m2g32hKvxYY1NJASOrpop73zhO-Ur0CfpYIDr>

Sent from my IPhone

On Mar 1, 2013, at 4:13 PM, Arni Sumarlidason <Arn...@md...> wrote:

> Users,
>  
> I have 20 nodes sitting on a disk arrays, with multiple LUNs. when I issue queries – `select count(1) from table` for example, I am experiencing heavy writes and heavy reads. I expected the reads but not the writes and it has really thrown a wrench in the caching. My first assumption would be the log files, do you have any other ideas what could be causing all these writes with a select?

It could be vacuum. Also, did you just load data? Hint bits get updated, dirtying pages and causing them to be written.

>  
> Arni Sumarlidason | Software Engineer, Information Technology
> MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
> O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
> arn...@md...| http://www.mdaus.com 
>  
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general

Hi Arni,

Am guessing that hint bits could be causing the I/O. See here:

http://wiki.postgresql.org/wiki/Hint_Bits

Regards,
Nikhils

On Fri, Mar 1, 2013 at 8:43 PM, Arni Sumarlidason
<Arn...@md...> wrote:
> Users,
>
>
>
> I have 20 nodes sitting on a disk arrays, with multiple LUNs. when I issue
> queries – `select count(1) from table` for example, I am experiencing heavy
> writes and heavy reads. I expected the reads but not the writes and it has
> really thrown a wrench in the caching. My first assumption would be the log
> files, do you have any other ideas what could be causing all these writes
> with a select?
>
>
>
> Arni Sumarlidason | Software Engineer, Information Technology
>
> MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
>
> O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
>
> arn...@md...| http://www.mdaus.com
>
>
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_feb
> _______________________________________________
> Postgres-xc-general mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>



-- 
StormDB - http://www.stormdb.com
The Database Cloud
Postgres-XC Support and Service

Users,

I have 20 nodes sitting on a disk arrays, with multiple LUNs. when I issue queries - `select count(1) from table` for example, I am experiencing heavy writes and heavy reads. I expected the reads but not the writes and it has really thrown a wrench in the caching. My first assumption would be the log files, do you have any other ideas what could be causing all these writes with a select?

Arni Sumarlidason | Software Engineer, Information Technology
MDA | 820 West Diamond Ave | Gaithersburg, MD | USA
O: 240-833-8200 D: 240-833-8318 M: 256-393-2803
arn...@md...<mailto:arn...@md...>| http://www.mdaus.com<http://www.mdaus.com/>

I see.   Maybe we should have your experience in our Wiki page for other's.

Regards;
----------
Koichi Suzuki


2013/3/1 Maz Mohammadi <mmo...@pe...>:
> Hi Koichi,
>
> Thanks for replying to my email.  http://www.postgresql.org/docs/9.2/static/ssl-tcp.html  is a good introductory document.  But I already got the SSL working on postgres-xc and verified it from psql.  The best document to set it up was http://gurjeet-tech.blogspot.in/2012/04/generating-self-signed-ssl-certificates.html
>
> My problem now is connecting postgres from a java application server (tomcat).  I've already read and followed the instructions in http://jdbc.postgresql.org/documentation/91/ssl-client.html .  Frankly there are on 2 steps....
>
> 1) openssl x509 -in pg_server.crt -out pg_server.crt.der -outform der
> 2) keytool -keystore $JAVA_HOME/lib/security/cacerts -alias postgresql -import -file pg_server.crt.der
>
> I also augmented the JDBC url with the right parameters.
> ssl=true and
> sslfactory=org.postgresql.ssl.NonValidatingFactory
> I've added the JVM parameters....
> java -Djavax.net.ssl.trustStore=mystore -Djavax.net.ssl.trustStorePassword=password
>
> something I want to mention is that postgres-xc is running as user "postgres-xc".  My tomcat is running as user "adminuser"
>
> anything I'm missing on the java side?
>
> It looks like no one has done this before. :(
>
> -maz
>
> -----Original Message-----
> From: Koichi Suzuki [mailto:koi...@gm...]
> Sent: Thursday, February 28, 2013 2:58 AM
> To: Maz Mohammadi
> Cc: Jim Mlodgenski; pos...@li...
> Subject: Re: [Postgres-xc-general] can't access server through SSL
>
> How's
> http://www.postgresql.org/docs/9.2/static/ssl-tcp.html ?
> https://kb.berkeley.edu/page.php?id=23113 may include some more info.
> ----------
> Koichi Suzuki
>
>
> 2013/2/27 Maz Mohammadi <mmo...@pe...>:
>> BTW, I also added the mapping option in pg_hba.conf and created a
>> mapping in pg_ident.conf which maps the OS user that runs tomcat to my
>> postgres-xc user.  Still I'm getting the same error :(
>>
>> I've also turned on the debugging of ssl on jvm and getting a lot of message on console?  Can anyone suggest on what I should look for?
>>
>> -maz
>>
>> -----Original Message-----
>> From: Maz Mohammadi
>> Sent: Tuesday, February 26, 2013 9:18 AM
>> To: 'Jim Mlodgenski'
>> Cc: Nikhil Sontakke; pos...@li...
>> Subject: RE: [Postgres-xc-general] can't access server through SSL
>>
>> Well,
>>
>> First let me say, this document http://gurjeet-tech.blogspot.in/2012/04/generating-self-signed-ssl-certificates.html  was very helpful in enabling SSL on my server + accessing it through psql using a certificate.  It clearly specifies how each file (server.* and client.*) is created and how it is used.
>>
>> As for the 2nd part my problem of reaching it from tomcat server, I'm still getting the same error.
>> "Connection attempt failed: FATAL: connection requires a valid client certificate"
>>
>> http://jdbc.postgresql.org/documentation/91/ssl-client.html#nonvalidating does not contain a very complex set of steps.
>>
>> 1) openssl x509 -in pg_server.crt -out pg_server.crt.der -outform der
>> 2) keytool -keystore $JAVA_HOME/lib/security/cacerts -alias postgresql
>> -import -file pg_server.crt.der
>>
>> pg_server.crt is the same as server.crt file that I got from the blog posting's instructions.
>>
>> And this is my url.
>> "jdbc:postgresql://localhost:5432:testdb?ssl=true&sslfactory=org.postgresql.ssl.NonValidatingFactory"
>>
>> Do I need to sign the certificate?
>>
>> -----Original Message-----
>> From: Jim Mlodgenski [mailto:ji...@gm...]
>> Sent: Monday, February 25, 2013 2:18 PM
>> To: Maz Mohammadi
>> Cc: Nikhil Sontakke; pos...@li...
>> Subject: Re: [Postgres-xc-general] can't access server through SSL
>>
>> On Mon, Feb 25, 2013 at 1:42 PM, Maz Mohammadi <mmo...@pe...> wrote:
>>> Tada....
>>>
>>> postgres-xc@adminuser-VirtualBox:~/.postgresql$ psql -h localhost -p
>>> 5432 testdb psql (PGXC 1.0.0, based on PG 9.1.4) SSL connection
>>> (cipher: DHE-RSA-AES256-SHA, bits: 256) Type "help" for help.
>>>
>>> testdb=#
>>>
>>> now I gotta get it working from tomcat ;)
>>
>> Take a look at
>> http://jdbc.postgresql.org/documentation/91/ssl-client.html#nonvalidat
>> ing
>>
>> You'll probably want to have JDBC not validate the cert against the CA
>> until you can at least connect from Tomcat. To do that, you'll need to
>> append the following to you connection string
>> sslfactory=org.postgresql.ssl.NonValidatingFactory
>>
>>>
>>> -maz
>>>
>>> -----Original Message-----
>>> From: Jim Mlodgenski [mailto:ji...@gm...]
>>> Sent: Monday, February 25, 2013 11:31 AM
>>> To: Maz Mohammadi
>>> Cc: Nikhil Sontakke; pos...@li...
>>> Subject: Re: [Postgres-xc-general] can't access server through SSL
>>>
>>> On Mon, Feb 25, 2013 at 11:12 AM, Maz Mohammadi <mmo...@pe...> wrote:
>>>> Something tells me, I might not be doing things right :(
>>>
>>> It does appear things are a little twisted up.
>>>
>>>>
>>>> postgres-xc@adminuser-VirtualBox:~/coord$ echo $PGSSLKEY
>>>> /var/lib/postgres-xc/.postgresql/client.key
>>>> postgres-xc@adminuser-VirtualBox:~/coord$ echo $PGSSLCERT
>>>> /var/lib/postgres-xc/coord/server.crt
>>>> postgres-xc@adminuser-VirtualBox:~/coord$ openssl verify -CAfile
>>>> ../coord/root.crt client.crt
>>>> client.crt: OK
>>>> postgres-xc@adminuser-VirtualBox:~/coord$ psql -U postgres-xc -h
>>>> localhost -p 5432
>>>> psql: could not load private key file
>>>> "/var/lib/postgres-xc/.postgresql/client.key": key values mismatch
>>>> postgres-xc@adminuser-VirtualBox:~/coord$
>>>>
>>>
>>> Start first with just making sure you can connect via SSL on the server before adding in the certificate authentication. Try changing your pg_hba to:
>>> hostssl all             postgres-xc             127.0.0.1/32            trust
>>>
>>> And see if you can connect via psql. There will be a message displayed when connected that it is an ssl connection.
>>>
>>> Once you get past that, you can revert back to the original pg_hba and focus on the cert auth. Getting the client certificate correct is a little tricky. I believe you need to sign it using the server certificate, but I need to look that up to be sure. I think you also need to make sure you have the proper mapping in the pg_ident file even though the names might be the same.
>>>
>>>> when I generated the key for the client, I used 'postgres-xc' for Common Name, and when I generated it for the server, I used 'localhost'
>>>>
>>>> Do you think I'm digging myself into a whole?  Should I start from scratch and install postgres?
>>>
>>> You'll have the same issue using PostgreSQL. Its not a XC vs PG issue, just a severe lack of documentation on how to do it properly.
>>>
>>>>
>>>> -maz
>>>>
>>>> -----Original Message-----
>>>> From: Nikhil Sontakke [mailto:ni...@st...]
>>>> Sent: Monday, February 25, 2013 10:21 AM
>>>> To: Maz Mohammadi
>>>> Cc: Michael Paquier; pos...@li...
>>>> Subject: Re: [Postgres-xc-general] can't access server through SSL
>>>>
>>>> Try using
>>>>
>>>> PGSSLKEY=/path/to/agent.key in psql.
>>>>
>>>> Regards,
>>>> Nikhils
>>>>
>>>> On Mon, Feb 25, 2013 at 7:51 PM, Maz Mohammadi <mmo...@pe...> wrote:
>>>>> Well,  it seems that this is the way to connect it from the
>>>>> standard psql client.  But I'm getting there ;(
>>>>>
>>>>>
>>>>>
>>>>> --------------
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ whoami
>>>>>
>>>>> postgres-xc
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ echo $PSGSSLMODE
>>>>>
>>>>> require
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ echo $PGSSLCERT
>>>>>
>>>>> /var/lib/postgres-xc/datanode1/server.crt
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ ls -l server.*
>>>>>
>>>>> -rw-rw-r-- 1 postgres-xc postgres-xc 4608 Feb 25 09:00 server.crt
>>>>>
>>>>> -rw------- 1 postgres-xc postgres-xc 1679 Feb 25 09:00 server.key
>>>>>
>>>>> -rw-rw-r-- 1 postgres-xc postgres-xc 3587 Feb 25 09:00 server.req
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ psql -U postgres-xc -h
>>>>> localhost -p 5432
>>>>>
>>>>> psql: certificate present, but not private key file
>>>>> "/var/lib/postgres-xc/.postgresql/postgresql.key"
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$ pwd
>>>>>
>>>>> /var/lib/postgres-xc/coord
>>>>>
>>>>> postgres-xc@adminuser-VirtualBox:~/coord$
>>>>>
>>>>> --------------
>>>>>
>>>>>
>>>>>
>>>>> Does the user 'postgres-xc' need to generate a private key file?
>>>>>
>>>>>
>>>>>
>>>>> postgresql.key does not exist.  I take it that it IS NOT the same
>>>>> as server.key which is under /var/lib/postgres-xc/coord.  Am I correct?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> From: Michael Paquier [mailto:mic...@gm...]
>>>>> Sent: Monday, February 25, 2013 2:17 AM
>>>>> To: Nikhil Sontakke
>>>>> Cc: Maz Mohammadi; pos...@li...
>>>>>
>>>>>
>>>>> Subject: Re: [Postgres-xc-general] can't access server through SSL
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Mon, Feb 25, 2013 at 4:10 PM, Nikhil Sontakke
>>>>> <ni...@st...>
>>>>> wrote:
>>>>>
>>>>> Hi Maz,
>>>>>
>>>>>
>>>>>> Is there a way to connect to the server using the psql by using a
>>>>>> certificate?
>>>>>>
>>>>>
>>>>> Does the below work for you?
>>>>>
>>>>> PSGSSLMODE=require PGSSLCERT=/path/to/agent.crt psql -d postgres -h
>>>>> localhost -p 5432
>>>>>
>>>>> Change PSGSSLMODE by PGSSLMODE. My 2c.
>>>>>
>>>>> --
>>>>> Michael
>>>>
>>>>
>>>>
>>>> --
>>>> StormDB - http://www.stormdb.com
>>>> The Database Cloud
>>>> Postgres-XC Support and Service
>>>>
>>>> --------------------------------------------------------------------
>>>> -
>>>> -
>>>> -------- Everyone hates slow websites. So do we.
>>>> Make your web apps faster with AppDynamics Download AppDynamics Lite
>>>> for free today:
>>>> http://p.sf.net/sfu/appdyn_d2d_feb
>>>> _______________________________________________
>>>> Postgres-xc-general mailing list
>>>> Pos...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general
>>
>> ----------------------------------------------------------------------
>> -------- Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics Download AppDynamics Lite
>> for free today:
>> http://p.sf.net/sfu/appdyn_d2d_feb
>> _______________________________________________
>> Postgres-xc-general mailing list
>> Pos...@li...
>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general