Menu
▾
▴
postgres-xc-general — General info and messages
You can subscribe to this list here.
| 2010 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
(6) |
Sep
|
Oct
(19) |
Nov
(1) |
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2011 |
Jan
(12) |
Feb
(1) |
Mar
(4) |
Apr
(4) |
May
(32) |
Jun
(12) |
Jul
(11) |
Aug
(1) |
Sep
(6) |
Oct
(3) |
Nov
|
Dec
(10) |
| 2012 |
Jan
(11) |
Feb
(1) |
Mar
(3) |
Apr
(25) |
May
(53) |
Jun
(38) |
Jul
(103) |
Aug
(54) |
Sep
(31) |
Oct
(66) |
Nov
(77) |
Dec
(20) |
| 2013 |
Jan
(91) |
Feb
(86) |
Mar
(103) |
Apr
(107) |
May
(25) |
Jun
(37) |
Jul
(17) |
Aug
(59) |
Sep
(38) |
Oct
(78) |
Nov
(29) |
Dec
(15) |
| 2014 |
Jan
(23) |
Feb
(82) |
Mar
(118) |
Apr
(101) |
May
(103) |
Jun
(45) |
Jul
(6) |
Aug
(10) |
Sep
|
Oct
(32) |
Nov
|
Dec
(9) |
| 2015 |
Jan
(3) |
Feb
(5) |
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
(9) |
Aug
(4) |
Sep
(3) |
Oct
|
Nov
|
Dec
|
| 2016 |
Jan
(3) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2017 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(3) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2018 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(4) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| S | M | T | W | T | F | S |
|---|---|---|---|---|---|---|
|
|
|
|
1
(2) |
2
(1) |
3
|
4
|
|
5
|
6
|
7
|
8
(1) |
9
(1) |
10
|
11
|
|
12
(1) |
13
(5) |
14
|
15
|
16
|
17
|
18
|
|
19
|
20
|
21
|
22
(4) |
23
(10) |
24
|
25
|
|
26
|
27
|
28
|
29
|
30
|
31
|
|
|
From: Donald K. <don...@gm...> - 2013-05-23 21:05:28
|
OMG...i am such a dunce! I didn't realize i put on subnet ... sorry about that. Thank you! On Thu, May 23, 2013 at 4:55 PM, Andrei Martsinchyk < and...@gm...> wrote: > Specify exact IP addresses of the coordinators, not subnets: > > host all all 100.100.10.170/*32* trust > host all all 100.100.10.171/*32* trust > > this is secure enough. > You do not have to enable connections from GTM, it never connects to > nodes, just accepts connections. You should just allow connections from > other coordinators. > So, if you have two coordinators 100.100.10.170 and 100.100.10.171, you > should put > host all all 100.100.10.171/32 trust > to pg_hba.conf of coordinator running on 100.100.10.170, and > host all all 100.100.10.170/32 trust > to pg_hba.conf of coordinator running on 100.100.10.171. > All datanodes need both these lines. > > > > > 2013/5/23 Donald Keyass <don...@gm...> > >> Right i did that ... by putting it for the coordinator servers. >> >> # IPv4 local connections: >> host all all 100.100.10.170/24 trust >> host all all 100.100.10.171/24 trust >> host all all 100.100.10.178/24 trust >> #GTM >> host all all 127.0.0.1/32 md5 >> host all all 100.100.10.0/24 md5 >> >> however, that means there's no password auth occuring at all then? B/c i >> then tried to connect from another server (a 9.2 db instance which has the >> client) and another server which had 9.0, I did: >> >> psql -h 10.10.0.170 -d testdb -U testuser >> >> No password prompt. and no, the pgpass file does not have an entry for >> these coordinator servers. >> >> So in this case that means i'm essentially making the PG-XC instance open >> to users w/o password auth. When I changed to md5, failed. >> >> >> >> >> On Thu, May 23, 2013 at 4:05 PM, Andrei Martsinchyk < >> and...@gm...> wrote: >> >>> Ah, sorry, I forgot to mention that coordinators should also be >>> configured to accept connections from other coordinators without a password. >>> Coordinator is sending DDLs to datanodes and coordinators, while DMLs >>> and SELECTs go only to datanodes. >>> That's why you are getting the error when trying to execute a DDL. >>> >>> >>> 2013/5/23 Donald Keyass <don...@gm...> >>> >>>> Alright i should apologize. I can do DMLs, but I can not do ddls. it >>>> seems only the postgres sys user can do that. I created a new postgres >>>> user w/ superuser privs: >>>> >>>> CREATE ROLE anotheruser WITH LOGIN PASSWORD 'password123' SUPERUSER >>>> INHERIT CREATEDB CREATEROLE; >>>> >>>> Granting privs, etc... >>>> Still can't do ddl's. i get the >>>> "ERROR: Failed to get pooled connections". >>>> >>>> when i create the user ... i do it through the coordinator. should I >>>> be creating user through the datanode on each datanode? >>>> >>>> >>>> On Thu, May 23, 2013 at 11:21 AM, Donald Keyass < >>>> don...@gm...> wrote: >>>> >>>>> actually selecting for the tables works. So I can run queries. But i >>>>> can't do any dmls or ddls in the database (ie create tables, >>>>> insert/update/delete). >>>>> >>>>> >>>>> On Thu, May 23, 2013 at 10:59 AM, Donald Keyass < >>>>> don...@gm...> wrote: >>>>> >>>>>> Andrei, >>>>>> Thanks! That was it. I set the hba for the datanodes as the same as >>>>>> coordinators. There was no documentation that I saw about the datanodes >>>>>> hba has to be passwordless (or in trust mode from all coordinators for all >>>>>> users). >>>>>> Again, thanks for the help! >>>>>> >>>>>> >>>>>> On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < >>>>>> and...@gm...> wrote: >>>>>> >>>>>>> Hi Donald, >>>>>>> >>>>>>> Please checke your pg_hba.conf on the datanodes, they should allow >>>>>>> passwordless access from coordinators. >>>>>>> When coordinator builds up connection string it does not include the >>>>>>> password. >>>>>>> >>>>>>> >>>>>>> 2013/5/23 Koichi Suzuki <koi...@gm...> >>>>>>> >>>>>>>> The following is a typical additional postgresql.conf file lines >>>>>>>> for coordinators. They're generated by pgxc_ctl utility and works well. >>>>>>>> Could you compare this with your ones? Especially, I'm interested how you >>>>>>>> configured pooler_port. >>>>>>>> >>>>>>>> Regards; >>>>>>>> >>>>>>>> ----8<----8<---- >>>>>>>> #=========================================== >>>>>>>> # Added at initialization. 20130523_11:33:33 >>>>>>>> port = 20004 >>>>>>>> pooler_port = 20010 >>>>>>>> gtm_host = 'node13' >>>>>>>> gtm_port = 20001 >>>>>>>> # End of Additon >>>>>>>> ---->8------->8------ >>>>>>>> >>>>>>>> >>>>>>>> ---------- >>>>>>>> Koichi Suzuki >>>>>>>> >>>>>>>> >>>>>>>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>>>>>>> >>>>>>>>> Hi Donald, >>>>>>>>> >>>>>>>>> This looks like a user permissions related internal bug in PGXC. >>>>>>>>> >>>>>>>>> Regards, >>>>>>>>> Nikhils >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>>>>>>> don...@gm...> wrote: >>>>>>>>> >>>>>>>>>> The testdb and user were the very last thing I created after I >>>>>>>>>> created the node. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>>>>>>> koi...@gm...> wrote: >>>>>>>>>> >>>>>>>>>>> Did you issue CREATE NODE before you create databases/roles? >>>>>>>>>>> If you CREATE NODE after you create databases/users, they are not visible >>>>>>>>>>> to other node and may fall into errors. When you create each node by >>>>>>>>>>> initdb, you also create gtm's using initgtm, connect all the >>>>>>>>>>> datanodes/coordinators to the gtm and then issue CREATE NODE before you >>>>>>>>>>> create anothing else. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> You can visit >>>>>>>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>>>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>>>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>>>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>>>>>>> Postgres-XC feature. >>>>>>>>>>> >>>>>>>>>>> Regards; >>>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>>>>> New Relic is the only SaaS-based application performance >>>>>>>>>> monitoring service >>>>>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>>>>> your >>>>>>>>>> browser, app, & servers with just a few lines of code. Try New >>>>>>>>>> Relic >>>>>>>>>> and get this awesome Nerd Life shirt! >>>>>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>>>>> _______________________________________________ >>>>>>>>>> Postgres-xc-general mailing list >>>>>>>>>> Pos...@li... >>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> -- >>>>>>>>> StormDB - http://www.stormdb.com >>>>>>>>> The Database Cloud >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> ------------------------------------------------------------------------------ >>>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>>>> service >>>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>>> your >>>>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>>>> and get this awesome Nerd Life shirt! >>>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>>> _______________________________________________ >>>>>>>> Postgres-xc-general mailing list >>>>>>>> Pos...@li... >>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Andrei Martsinchyk >>>>>>> >>>>>>> >>>>>>> StormDB - http://www.stormdb.com >>>>>>> The Database Cloud >>>>>>> >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------------------------ >>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>>> service >>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>> your >>>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>>> and get this awesome Nerd Life shirt! >>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>> _______________________________________________ >>>>>>> Postgres-xc-general mailing list >>>>>>> Pos...@li... >>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >>> >>> -- >>> Andrei Martsinchyk >>> >>> StormDB - http://www.stormdb.com >>> The Database Cloud >>> >>> >> > > > -- > Andrei Martsinchyk > > StormDB - http://www.stormdb.com > The Database Cloud > > |
|
From: Andrei M. <and...@gm...> - 2013-05-23 20:55:12
|
Specify exact IP addresses of the coordinators, not subnets: host all all 100.100.10.170/*32* trust host all all 100.100.10.171/*32* trust this is secure enough. You do not have to enable connections from GTM, it never connects to nodes, just accepts connections. You should just allow connections from other coordinators. So, if you have two coordinators 100.100.10.170 and 100.100.10.171, you should put host all all 100.100.10.171/32 trust to pg_hba.conf of coordinator running on 100.100.10.170, and host all all 100.100.10.170/32 trust to pg_hba.conf of coordinator running on 100.100.10.171. All datanodes need both these lines. 2013/5/23 Donald Keyass <don...@gm...> > Right i did that ... by putting it for the coordinator servers. > > # IPv4 local connections: > host all all 100.100.10.170/24 trust > host all all 100.100.10.171/24 trust > host all all 100.100.10.178/24 trust > #GTM > host all all 127.0.0.1/32 md5 > host all all 100.100.10.0/24 md5 > > however, that means there's no password auth occuring at all then? B/c i > then tried to connect from another server (a 9.2 db instance which has the > client) and another server which had 9.0, I did: > > psql -h 10.10.0.170 -d testdb -U testuser > > No password prompt. and no, the pgpass file does not have an entry for > these coordinator servers. > > So in this case that means i'm essentially making the PG-XC instance open > to users w/o password auth. When I changed to md5, failed. > > > > > On Thu, May 23, 2013 at 4:05 PM, Andrei Martsinchyk < > and...@gm...> wrote: > >> Ah, sorry, I forgot to mention that coordinators should also be >> configured to accept connections from other coordinators without a password. >> Coordinator is sending DDLs to datanodes and coordinators, while DMLs and >> SELECTs go only to datanodes. >> That's why you are getting the error when trying to execute a DDL. >> >> >> 2013/5/23 Donald Keyass <don...@gm...> >> >>> Alright i should apologize. I can do DMLs, but I can not do ddls. it >>> seems only the postgres sys user can do that. I created a new postgres >>> user w/ superuser privs: >>> >>> CREATE ROLE anotheruser WITH LOGIN PASSWORD 'password123' SUPERUSER >>> INHERIT CREATEDB CREATEROLE; >>> >>> Granting privs, etc... >>> Still can't do ddl's. i get the >>> "ERROR: Failed to get pooled connections". >>> >>> when i create the user ... i do it through the coordinator. should I >>> be creating user through the datanode on each datanode? >>> >>> >>> On Thu, May 23, 2013 at 11:21 AM, Donald Keyass < >>> don...@gm...> wrote: >>> >>>> actually selecting for the tables works. So I can run queries. But i >>>> can't do any dmls or ddls in the database (ie create tables, >>>> insert/update/delete). >>>> >>>> >>>> On Thu, May 23, 2013 at 10:59 AM, Donald Keyass < >>>> don...@gm...> wrote: >>>> >>>>> Andrei, >>>>> Thanks! That was it. I set the hba for the datanodes as the same as >>>>> coordinators. There was no documentation that I saw about the datanodes >>>>> hba has to be passwordless (or in trust mode from all coordinators for all >>>>> users). >>>>> Again, thanks for the help! >>>>> >>>>> >>>>> On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < >>>>> and...@gm...> wrote: >>>>> >>>>>> Hi Donald, >>>>>> >>>>>> Please checke your pg_hba.conf on the datanodes, they should allow >>>>>> passwordless access from coordinators. >>>>>> When coordinator builds up connection string it does not include the >>>>>> password. >>>>>> >>>>>> >>>>>> 2013/5/23 Koichi Suzuki <koi...@gm...> >>>>>> >>>>>>> The following is a typical additional postgresql.conf file lines for >>>>>>> coordinators. They're generated by pgxc_ctl utility and works well. >>>>>>> Could you compare this with your ones? Especially, I'm interested how you >>>>>>> configured pooler_port. >>>>>>> >>>>>>> Regards; >>>>>>> >>>>>>> ----8<----8<---- >>>>>>> #=========================================== >>>>>>> # Added at initialization. 20130523_11:33:33 >>>>>>> port = 20004 >>>>>>> pooler_port = 20010 >>>>>>> gtm_host = 'node13' >>>>>>> gtm_port = 20001 >>>>>>> # End of Additon >>>>>>> ---->8------->8------ >>>>>>> >>>>>>> >>>>>>> ---------- >>>>>>> Koichi Suzuki >>>>>>> >>>>>>> >>>>>>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>>>>>> >>>>>>>> Hi Donald, >>>>>>>> >>>>>>>> This looks like a user permissions related internal bug in PGXC. >>>>>>>> >>>>>>>> Regards, >>>>>>>> Nikhils >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>>>>>> don...@gm...> wrote: >>>>>>>> >>>>>>>>> The testdb and user were the very last thing I created after I >>>>>>>>> created the node. >>>>>>>>> >>>>>>>>> >>>>>>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>>>>>> koi...@gm...> wrote: >>>>>>>>> >>>>>>>>>> Did you issue CREATE NODE before you create databases/roles? If >>>>>>>>>> you CREATE NODE after you create databases/users, they are not visible to >>>>>>>>>> other node and may fall into errors. When you create each node by initdb, >>>>>>>>>> you also create gtm's using initgtm, connect all the datanodes/coordinators >>>>>>>>>> to the gtm and then issue CREATE NODE before you create anothing else. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> You can visit >>>>>>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>>>>>> Postgres-XC feature. >>>>>>>>>> >>>>>>>>>> Regards; >>>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>>>> New Relic is the only SaaS-based application performance >>>>>>>>> monitoring service >>>>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>>>> your >>>>>>>>> browser, app, & servers with just a few lines of code. Try New >>>>>>>>> Relic >>>>>>>>> and get this awesome Nerd Life shirt! >>>>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>>>> _______________________________________________ >>>>>>>>> Postgres-xc-general mailing list >>>>>>>>> Pos...@li... >>>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> StormDB - http://www.stormdb.com >>>>>>>> The Database Cloud >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------------------------ >>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>>> service >>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>> your >>>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>>> and get this awesome Nerd Life shirt! >>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>> _______________________________________________ >>>>>>> Postgres-xc-general mailing list >>>>>>> Pos...@li... >>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> Andrei Martsinchyk >>>>>> >>>>>> >>>>>> StormDB - http://www.stormdb.com >>>>>> The Database Cloud >>>>>> >>>>>> >>>>>> >>>>>> ------------------------------------------------------------------------------ >>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>> service >>>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>> and get this awesome Nerd Life shirt! >>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>> _______________________________________________ >>>>>> Postgres-xc-general mailing list >>>>>> Pos...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>> >>>>>> >>>>> >>>> >>> >> >> >> -- >> Andrei Martsinchyk >> >> StormDB - http://www.stormdb.com >> The Database Cloud >> >> > -- Andrei Martsinchyk StormDB - http://www.stormdb.com The Database Cloud |
|
From: Donald K. <don...@gm...> - 2013-05-23 20:14:06
|
Right i did that ... by putting it for the coordinator servers. # IPv4 local connections: host all all 100.100.10.170/24 trust host all all 100.100.10.171/24 trust host all all 100.100.10.178/24 trust #GTM host all all 127.0.0.1/32 md5 host all all 100.100.10.0/24 md5 however, that means there's no password auth occuring at all then? B/c i then tried to connect from another server (a 9.2 db instance which has the client) and another server which had 9.0, I did: psql -h 10.10.0.170 -d testdb -U testuser No password prompt. and no, the pgpass file does not have an entry for these coordinator servers. So in this case that means i'm essentially making the PG-XC instance open to users w/o password auth. When I changed to md5, failed. On Thu, May 23, 2013 at 4:05 PM, Andrei Martsinchyk < and...@gm...> wrote: > Ah, sorry, I forgot to mention that coordinators should also be configured > to accept connections from other coordinators without a password. > Coordinator is sending DDLs to datanodes and coordinators, while DMLs and > SELECTs go only to datanodes. > That's why you are getting the error when trying to execute a DDL. > > > 2013/5/23 Donald Keyass <don...@gm...> > >> Alright i should apologize. I can do DMLs, but I can not do ddls. it >> seems only the postgres sys user can do that. I created a new postgres >> user w/ superuser privs: >> >> CREATE ROLE anotheruser WITH LOGIN PASSWORD 'password123' SUPERUSER >> INHERIT CREATEDB CREATEROLE; >> >> Granting privs, etc... >> Still can't do ddl's. i get the >> "ERROR: Failed to get pooled connections". >> >> when i create the user ... i do it through the coordinator. should I be >> creating user through the datanode on each datanode? >> >> >> On Thu, May 23, 2013 at 11:21 AM, Donald Keyass < >> don...@gm...> wrote: >> >>> actually selecting for the tables works. So I can run queries. But i >>> can't do any dmls or ddls in the database (ie create tables, >>> insert/update/delete). >>> >>> >>> On Thu, May 23, 2013 at 10:59 AM, Donald Keyass < >>> don...@gm...> wrote: >>> >>>> Andrei, >>>> Thanks! That was it. I set the hba for the datanodes as the same as >>>> coordinators. There was no documentation that I saw about the datanodes >>>> hba has to be passwordless (or in trust mode from all coordinators for all >>>> users). >>>> Again, thanks for the help! >>>> >>>> >>>> On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < >>>> and...@gm...> wrote: >>>> >>>>> Hi Donald, >>>>> >>>>> Please checke your pg_hba.conf on the datanodes, they should allow >>>>> passwordless access from coordinators. >>>>> When coordinator builds up connection string it does not include the >>>>> password. >>>>> >>>>> >>>>> 2013/5/23 Koichi Suzuki <koi...@gm...> >>>>> >>>>>> The following is a typical additional postgresql.conf file lines for >>>>>> coordinators. They're generated by pgxc_ctl utility and works well. >>>>>> Could you compare this with your ones? Especially, I'm interested how you >>>>>> configured pooler_port. >>>>>> >>>>>> Regards; >>>>>> >>>>>> ----8<----8<---- >>>>>> #=========================================== >>>>>> # Added at initialization. 20130523_11:33:33 >>>>>> port = 20004 >>>>>> pooler_port = 20010 >>>>>> gtm_host = 'node13' >>>>>> gtm_port = 20001 >>>>>> # End of Additon >>>>>> ---->8------->8------ >>>>>> >>>>>> >>>>>> ---------- >>>>>> Koichi Suzuki >>>>>> >>>>>> >>>>>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>>>>> >>>>>>> Hi Donald, >>>>>>> >>>>>>> This looks like a user permissions related internal bug in PGXC. >>>>>>> >>>>>>> Regards, >>>>>>> Nikhils >>>>>>> >>>>>>> >>>>>>> >>>>>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>>>>> don...@gm...> wrote: >>>>>>> >>>>>>>> The testdb and user were the very last thing I created after I >>>>>>>> created the node. >>>>>>>> >>>>>>>> >>>>>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>>>>> koi...@gm...> wrote: >>>>>>>> >>>>>>>>> Did you issue CREATE NODE before you create databases/roles? If >>>>>>>>> you CREATE NODE after you create databases/users, they are not visible to >>>>>>>>> other node and may fall into errors. When you create each node by initdb, >>>>>>>>> you also create gtm's using initgtm, connect all the datanodes/coordinators >>>>>>>>> to the gtm and then issue CREATE NODE before you create anothing else. >>>>>>>>> >>>>>>>>> >>>>>>>>> You can visit >>>>>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>>>>> Postgres-XC feature. >>>>>>>>> >>>>>>>>> Regards; >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> ------------------------------------------------------------------------------ >>>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>>>> service >>>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>>> your >>>>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>>>> and get this awesome Nerd Life shirt! >>>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>>> _______________________________________________ >>>>>>>> Postgres-xc-general mailing list >>>>>>>> Pos...@li... >>>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> StormDB - http://www.stormdb.com >>>>>>> The Database Cloud >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> ------------------------------------------------------------------------------ >>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>> service >>>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>> and get this awesome Nerd Life shirt! >>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>> _______________________________________________ >>>>>> Postgres-xc-general mailing list >>>>>> Pos...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> Andrei Martsinchyk >>>>> >>>>> >>>>> StormDB - http://www.stormdb.com >>>>> The Database Cloud >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------------ >>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>> New Relic is the only SaaS-based application performance monitoring >>>>> service >>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>> and get this awesome Nerd Life shirt! >>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>> _______________________________________________ >>>>> Postgres-xc-general mailing list >>>>> Pos...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>> >>>>> >>>> >>> >> > > > -- > Andrei Martsinchyk > > StormDB - http://www.stormdb.com > The Database Cloud > > |
|
From: Andrei M. <and...@gm...> - 2013-05-23 20:06:08
|
Ah, sorry, I forgot to mention that coordinators should also be configured to accept connections from other coordinators without a password. Coordinator is sending DDLs to datanodes and coordinators, while DMLs and SELECTs go only to datanodes. That's why you are getting the error when trying to execute a DDL. 2013/5/23 Donald Keyass <don...@gm...> > Alright i should apologize. I can do DMLs, but I can not do ddls. it > seems only the postgres sys user can do that. I created a new postgres > user w/ superuser privs: > > CREATE ROLE anotheruser WITH LOGIN PASSWORD 'password123' SUPERUSER > INHERIT CREATEDB CREATEROLE; > > Granting privs, etc... > Still can't do ddl's. i get the > "ERROR: Failed to get pooled connections". > > when i create the user ... i do it through the coordinator. should I be > creating user through the datanode on each datanode? > > > On Thu, May 23, 2013 at 11:21 AM, Donald Keyass < > don...@gm...> wrote: > >> actually selecting for the tables works. So I can run queries. But i >> can't do any dmls or ddls in the database (ie create tables, >> insert/update/delete). >> >> >> On Thu, May 23, 2013 at 10:59 AM, Donald Keyass < >> don...@gm...> wrote: >> >>> Andrei, >>> Thanks! That was it. I set the hba for the datanodes as the same as >>> coordinators. There was no documentation that I saw about the datanodes >>> hba has to be passwordless (or in trust mode from all coordinators for all >>> users). >>> Again, thanks for the help! >>> >>> >>> On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < >>> and...@gm...> wrote: >>> >>>> Hi Donald, >>>> >>>> Please checke your pg_hba.conf on the datanodes, they should allow >>>> passwordless access from coordinators. >>>> When coordinator builds up connection string it does not include the >>>> password. >>>> >>>> >>>> 2013/5/23 Koichi Suzuki <koi...@gm...> >>>> >>>>> The following is a typical additional postgresql.conf file lines for >>>>> coordinators. They're generated by pgxc_ctl utility and works well. >>>>> Could you compare this with your ones? Especially, I'm interested how you >>>>> configured pooler_port. >>>>> >>>>> Regards; >>>>> >>>>> ----8<----8<---- >>>>> #=========================================== >>>>> # Added at initialization. 20130523_11:33:33 >>>>> port = 20004 >>>>> pooler_port = 20010 >>>>> gtm_host = 'node13' >>>>> gtm_port = 20001 >>>>> # End of Additon >>>>> ---->8------->8------ >>>>> >>>>> >>>>> ---------- >>>>> Koichi Suzuki >>>>> >>>>> >>>>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>>>> >>>>>> Hi Donald, >>>>>> >>>>>> This looks like a user permissions related internal bug in PGXC. >>>>>> >>>>>> Regards, >>>>>> Nikhils >>>>>> >>>>>> >>>>>> >>>>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>>>> don...@gm...> wrote: >>>>>> >>>>>>> The testdb and user were the very last thing I created after I >>>>>>> created the node. >>>>>>> >>>>>>> >>>>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>>>> koi...@gm...> wrote: >>>>>>> >>>>>>>> Did you issue CREATE NODE before you create databases/roles? If >>>>>>>> you CREATE NODE after you create databases/users, they are not visible to >>>>>>>> other node and may fall into errors. When you create each node by initdb, >>>>>>>> you also create gtm's using initgtm, connect all the datanodes/coordinators >>>>>>>> to the gtm and then issue CREATE NODE before you create anothing else. >>>>>>>> >>>>>>>> >>>>>>>> You can visit >>>>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>>>> Postgres-XC feature. >>>>>>>> >>>>>>>> Regards; >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------------------------ >>>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>>> service >>>>>>> that delivers powerful full stack analytics. Optimize and monitor >>>>>>> your >>>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>>> and get this awesome Nerd Life shirt! >>>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>>> _______________________________________________ >>>>>>> Postgres-xc-general mailing list >>>>>>> Pos...@li... >>>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> StormDB - http://www.stormdb.com >>>>>> The Database Cloud >>>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------------ >>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>> New Relic is the only SaaS-based application performance monitoring >>>>> service >>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>> and get this awesome Nerd Life shirt! >>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>> _______________________________________________ >>>>> Postgres-xc-general mailing list >>>>> Pos...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>> >>>>> >>>> >>>> >>>> -- >>>> Andrei Martsinchyk >>>> >>>> >>>> StormDB - http://www.stormdb.com >>>> The Database Cloud >>>> >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Try New Relic Now & We'll Send You this Cool Shirt >>>> New Relic is the only SaaS-based application performance monitoring >>>> service >>>> that delivers powerful full stack analytics. Optimize and monitor your >>>> browser, app, & servers with just a few lines of code. Try New Relic >>>> and get this awesome Nerd Life shirt! >>>> http://p.sf.net/sfu/newrelic_d2d_may >>>> _______________________________________________ >>>> Postgres-xc-general mailing list >>>> Pos...@li... >>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>> >>>> >>> >> > -- Andrei Martsinchyk StormDB - http://www.stormdb.com The Database Cloud |
|
From: Donald K. <don...@gm...> - 2013-05-23 19:16:00
|
Alright i should apologize. I can do DMLs, but I can not do ddls. it seems only the postgres sys user can do that. I created a new postgres user w/ superuser privs: CREATE ROLE anotheruser WITH LOGIN PASSWORD 'password123' SUPERUSER INHERIT CREATEDB CREATEROLE; Granting privs, etc... Still can't do ddl's. i get the "ERROR: Failed to get pooled connections". when i create the user ... i do it through the coordinator. should I be creating user through the datanode on each datanode? On Thu, May 23, 2013 at 11:21 AM, Donald Keyass <don...@gm...>wrote: > actually selecting for the tables works. So I can run queries. But i > can't do any dmls or ddls in the database (ie create tables, > insert/update/delete). > > > On Thu, May 23, 2013 at 10:59 AM, Donald Keyass < > don...@gm...> wrote: > >> Andrei, >> Thanks! That was it. I set the hba for the datanodes as the same as >> coordinators. There was no documentation that I saw about the datanodes >> hba has to be passwordless (or in trust mode from all coordinators for all >> users). >> Again, thanks for the help! >> >> >> On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < >> and...@gm...> wrote: >> >>> Hi Donald, >>> >>> Please checke your pg_hba.conf on the datanodes, they should allow >>> passwordless access from coordinators. >>> When coordinator builds up connection string it does not include the >>> password. >>> >>> >>> 2013/5/23 Koichi Suzuki <koi...@gm...> >>> >>>> The following is a typical additional postgresql.conf file lines for >>>> coordinators. They're generated by pgxc_ctl utility and works well. >>>> Could you compare this with your ones? Especially, I'm interested how you >>>> configured pooler_port. >>>> >>>> Regards; >>>> >>>> ----8<----8<---- >>>> #=========================================== >>>> # Added at initialization. 20130523_11:33:33 >>>> port = 20004 >>>> pooler_port = 20010 >>>> gtm_host = 'node13' >>>> gtm_port = 20001 >>>> # End of Additon >>>> ---->8------->8------ >>>> >>>> >>>> ---------- >>>> Koichi Suzuki >>>> >>>> >>>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>>> >>>>> Hi Donald, >>>>> >>>>> This looks like a user permissions related internal bug in PGXC. >>>>> >>>>> Regards, >>>>> Nikhils >>>>> >>>>> >>>>> >>>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>>> don...@gm...> wrote: >>>>> >>>>>> The testdb and user were the very last thing I created after I >>>>>> created the node. >>>>>> >>>>>> >>>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>>> koi...@gm...> wrote: >>>>>> >>>>>>> Did you issue CREATE NODE before you create databases/roles? If >>>>>>> you CREATE NODE after you create databases/users, they are not visible to >>>>>>> other node and may fall into errors. When you create each node by initdb, >>>>>>> you also create gtm's using initgtm, connect all the datanodes/coordinators >>>>>>> to the gtm and then issue CREATE NODE before you create anothing else. >>>>>>> >>>>>>> >>>>>>> You can visit >>>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>>> Postgres-XC feature. >>>>>>> >>>>>>> Regards; >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> ------------------------------------------------------------------------------ >>>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>>> New Relic is the only SaaS-based application performance monitoring >>>>>> service >>>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>>> and get this awesome Nerd Life shirt! >>>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>>> _______________________________________________ >>>>>> Postgres-xc-general mailing list >>>>>> Pos...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> StormDB - http://www.stormdb.com >>>>> The Database Cloud >>>>> >>>> >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Try New Relic Now & We'll Send You this Cool Shirt >>>> New Relic is the only SaaS-based application performance monitoring >>>> service >>>> that delivers powerful full stack analytics. Optimize and monitor your >>>> browser, app, & servers with just a few lines of code. Try New Relic >>>> and get this awesome Nerd Life shirt! >>>> http://p.sf.net/sfu/newrelic_d2d_may >>>> _______________________________________________ >>>> Postgres-xc-general mailing list >>>> Pos...@li... >>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>> >>>> >>> >>> >>> -- >>> Andrei Martsinchyk >>> >>> >>> StormDB - http://www.stormdb.com >>> The Database Cloud >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Try New Relic Now & We'll Send You this Cool Shirt >>> New Relic is the only SaaS-based application performance monitoring >>> service >>> that delivers powerful full stack analytics. Optimize and monitor your >>> browser, app, & servers with just a few lines of code. Try New Relic >>> and get this awesome Nerd Life shirt! >>> http://p.sf.net/sfu/newrelic_d2d_may >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >>> >> > |
|
From: Donald K. <don...@gm...> - 2013-05-23 15:21:32
|
actually selecting for the tables works. So I can run queries. But i can't do any dmls or ddls in the database (ie create tables, insert/update/delete). On Thu, May 23, 2013 at 10:59 AM, Donald Keyass <don...@gm...>wrote: > Andrei, > Thanks! That was it. I set the hba for the datanodes as the same as > coordinators. There was no documentation that I saw about the datanodes > hba has to be passwordless (or in trust mode from all coordinators for all > users). > Again, thanks for the help! > > > On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < > and...@gm...> wrote: > >> Hi Donald, >> >> Please checke your pg_hba.conf on the datanodes, they should allow >> passwordless access from coordinators. >> When coordinator builds up connection string it does not include the >> password. >> >> >> 2013/5/23 Koichi Suzuki <koi...@gm...> >> >>> The following is a typical additional postgresql.conf file lines for >>> coordinators. They're generated by pgxc_ctl utility and works well. >>> Could you compare this with your ones? Especially, I'm interested how you >>> configured pooler_port. >>> >>> Regards; >>> >>> ----8<----8<---- >>> #=========================================== >>> # Added at initialization. 20130523_11:33:33 >>> port = 20004 >>> pooler_port = 20010 >>> gtm_host = 'node13' >>> gtm_port = 20001 >>> # End of Additon >>> ---->8------->8------ >>> >>> >>> ---------- >>> Koichi Suzuki >>> >>> >>> 2013/5/22 Nikhil Sontakke <ni...@st...> >>> >>>> Hi Donald, >>>> >>>> This looks like a user permissions related internal bug in PGXC. >>>> >>>> Regards, >>>> Nikhils >>>> >>>> >>>> >>>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>>> don...@gm...> wrote: >>>> >>>>> The testdb and user were the very last thing I created after I created >>>>> the node. >>>>> >>>>> >>>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>>> koi...@gm...> wrote: >>>>> >>>>>> Did you issue CREATE NODE before you create databases/roles? If you >>>>>> CREATE NODE after you create databases/users, they are not visible to other >>>>>> node and may fall into errors. When you create each node by initdb, you >>>>>> also create gtm's using initgtm, connect all the datanodes/coordinators to >>>>>> the gtm and then issue CREATE NODE before you create anothing else. >>>>>> >>>>>> >>>>>> You can visit >>>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>>> failover, start and stop the cluster. init_all function will be helpful >>>>>> for you. This shows what you should do to initialize your cluster. The >>>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>>> Postgres-XC feature. >>>>>> >>>>>> Regards; >>>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------------ >>>>> Try New Relic Now & We'll Send You this Cool Shirt >>>>> New Relic is the only SaaS-based application performance monitoring >>>>> service >>>>> that delivers powerful full stack analytics. Optimize and monitor your >>>>> browser, app, & servers with just a few lines of code. Try New Relic >>>>> and get this awesome Nerd Life shirt! >>>>> http://p.sf.net/sfu/newrelic_d2d_may >>>>> _______________________________________________ >>>>> Postgres-xc-general mailing list >>>>> Pos...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>> >>>>> >>>> >>>> >>>> -- >>>> StormDB - http://www.stormdb.com >>>> The Database Cloud >>>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Try New Relic Now & We'll Send You this Cool Shirt >>> New Relic is the only SaaS-based application performance monitoring >>> service >>> that delivers powerful full stack analytics. Optimize and monitor your >>> browser, app, & servers with just a few lines of code. Try New Relic >>> and get this awesome Nerd Life shirt! >>> http://p.sf.net/sfu/newrelic_d2d_may >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >>> >> >> >> -- >> Andrei Martsinchyk >> >> >> StormDB - http://www.stormdb.com >> The Database Cloud >> >> >> >> ------------------------------------------------------------------------------ >> Try New Relic Now & We'll Send You this Cool Shirt >> New Relic is the only SaaS-based application performance monitoring >> service >> that delivers powerful full stack analytics. Optimize and monitor your >> browser, app, & servers with just a few lines of code. Try New Relic >> and get this awesome Nerd Life shirt! >> http://p.sf.net/sfu/newrelic_d2d_may >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >> >> > |
|
From: Donald K. <don...@gm...> - 2013-05-23 14:59:09
|
Andrei, Thanks! That was it. I set the hba for the datanodes as the same as coordinators. There was no documentation that I saw about the datanodes hba has to be passwordless (or in trust mode from all coordinators for all users). Again, thanks for the help! On Thu, May 23, 2013 at 3:36 AM, Andrei Martsinchyk < and...@gm...> wrote: > Hi Donald, > > Please checke your pg_hba.conf on the datanodes, they should allow > passwordless access from coordinators. > When coordinator builds up connection string it does not include the > password. > > > 2013/5/23 Koichi Suzuki <koi...@gm...> > >> The following is a typical additional postgresql.conf file lines for >> coordinators. They're generated by pgxc_ctl utility and works well. >> Could you compare this with your ones? Especially, I'm interested how you >> configured pooler_port. >> >> Regards; >> >> ----8<----8<---- >> #=========================================== >> # Added at initialization. 20130523_11:33:33 >> port = 20004 >> pooler_port = 20010 >> gtm_host = 'node13' >> gtm_port = 20001 >> # End of Additon >> ---->8------->8------ >> >> >> ---------- >> Koichi Suzuki >> >> >> 2013/5/22 Nikhil Sontakke <ni...@st...> >> >>> Hi Donald, >>> >>> This looks like a user permissions related internal bug in PGXC. >>> >>> Regards, >>> Nikhils >>> >>> >>> >>> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >>> don...@gm...> wrote: >>> >>>> The testdb and user were the very last thing I created after I created >>>> the node. >>>> >>>> >>>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>>> koi...@gm...> wrote: >>>> >>>>> Did you issue CREATE NODE before you create databases/roles? If you >>>>> CREATE NODE after you create databases/users, they are not visible to other >>>>> node and may fall into errors. When you create each node by initdb, you >>>>> also create gtm's using initgtm, connect all the datanodes/coordinators to >>>>> the gtm and then issue CREATE NODE before you create anothing else. >>>>> >>>>> >>>>> You can visit >>>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>>> failover, start and stop the cluster. init_all function will be helpful >>>>> for you. This shows what you should do to initialize your cluster. The >>>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>>> Postgres-XC feature. >>>>> >>>>> Regards; >>>>> >>>> >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Try New Relic Now & We'll Send You this Cool Shirt >>>> New Relic is the only SaaS-based application performance monitoring >>>> service >>>> that delivers powerful full stack analytics. Optimize and monitor your >>>> browser, app, & servers with just a few lines of code. Try New Relic >>>> and get this awesome Nerd Life shirt! >>>> http://p.sf.net/sfu/newrelic_d2d_may >>>> _______________________________________________ >>>> Postgres-xc-general mailing list >>>> Pos...@li... >>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>> >>>> >>> >>> >>> -- >>> StormDB - http://www.stormdb.com >>> The Database Cloud >>> >> >> >> >> ------------------------------------------------------------------------------ >> Try New Relic Now & We'll Send You this Cool Shirt >> New Relic is the only SaaS-based application performance monitoring >> service >> that delivers powerful full stack analytics. Optimize and monitor your >> browser, app, & servers with just a few lines of code. Try New Relic >> and get this awesome Nerd Life shirt! >> http://p.sf.net/sfu/newrelic_d2d_may >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >> >> > > > -- > Andrei Martsinchyk > > > StormDB - http://www.stormdb.com > The Database Cloud > > > > ------------------------------------------------------------------------------ > Try New Relic Now & We'll Send You this Cool Shirt > New Relic is the only SaaS-based application performance monitoring service > that delivers powerful full stack analytics. Optimize and monitor your > browser, app, & servers with just a few lines of code. Try New Relic > and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > |
|
From: Donald K. <don...@gm...> - 2013-05-23 14:50:22
|
I left the pooler port as the default which I believe is 6667. But I don't see how that effects a postgresql generated user vs the postgres user which can do ddls/dmls/and queries against tables. If the pooler port would effect pg generated user, wouldn't the postgres user as well be effected by this? On Wed, May 22, 2013 at 10:36 PM, Koichi Suzuki <koi...@gm...>wrote: > The following is a typical additional postgresql.conf file lines for > coordinators. They're generated by pgxc_ctl utility and works well. > Could you compare this with your ones? Especially, I'm interested how you > configured pooler_port. > > Regards; > > ----8<----8<---- > #=========================================== > # Added at initialization. 20130523_11:33:33 > port = 20004 > pooler_port = 20010 > gtm_host = 'node13' > gtm_port = 20001 > # End of Additon > ---->8------->8------ > > > ---------- > Koichi Suzuki > > > 2013/5/22 Nikhil Sontakke <ni...@st...> > >> Hi Donald, >> >> This looks like a user permissions related internal bug in PGXC. >> >> Regards, >> Nikhils >> >> >> >> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >> don...@gm...> wrote: >> >>> The testdb and user were the very last thing I created after I created >>> the node. >>> >>> >>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>> koi...@gm...> wrote: >>> >>>> Did you issue CREATE NODE before you create databases/roles? If you >>>> CREATE NODE after you create databases/users, they are not visible to other >>>> node and may fall into errors. When you create each node by initdb, you >>>> also create gtm's using initgtm, connect all the datanodes/coordinators to >>>> the gtm and then issue CREATE NODE before you create anothing else. >>>> >>>> >>>> You can visit >>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>> failover, start and stop the cluster. init_all function will be helpful >>>> for you. This shows what you should do to initialize your cluster. The >>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>> Postgres-XC feature. >>>> >>>> Regards; >>>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Try New Relic Now & We'll Send You this Cool Shirt >>> New Relic is the only SaaS-based application performance monitoring >>> service >>> that delivers powerful full stack analytics. Optimize and monitor your >>> browser, app, & servers with just a few lines of code. Try New Relic >>> and get this awesome Nerd Life shirt! >>> http://p.sf.net/sfu/newrelic_d2d_may >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >>> >> >> >> -- >> StormDB - http://www.stormdb.com >> The Database Cloud >> > > |
|
From: Andrei M. <and...@gm...> - 2013-05-23 07:36:58
|
Hi Donald, Please checke your pg_hba.conf on the datanodes, they should allow passwordless access from coordinators. When coordinator builds up connection string it does not include the password. 2013/5/23 Koichi Suzuki <koi...@gm...> > The following is a typical additional postgresql.conf file lines for > coordinators. They're generated by pgxc_ctl utility and works well. > Could you compare this with your ones? Especially, I'm interested how you > configured pooler_port. > > Regards; > > ----8<----8<---- > #=========================================== > # Added at initialization. 20130523_11:33:33 > port = 20004 > pooler_port = 20010 > gtm_host = 'node13' > gtm_port = 20001 > # End of Additon > ---->8------->8------ > > > ---------- > Koichi Suzuki > > > 2013/5/22 Nikhil Sontakke <ni...@st...> > >> Hi Donald, >> >> This looks like a user permissions related internal bug in PGXC. >> >> Regards, >> Nikhils >> >> >> >> On Thu, May 23, 2013 at 3:13 AM, Donald Keyass < >> don...@gm...> wrote: >> >>> The testdb and user were the very last thing I created after I created >>> the node. >>> >>> >>> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki < >>> koi...@gm...> wrote: >>> >>>> Did you issue CREATE NODE before you create databases/roles? If you >>>> CREATE NODE after you create databases/users, they are not visible to other >>>> node and may fall into errors. When you create each node by initdb, you >>>> also create gtm's using initgtm, connect all the datanodes/coordinators to >>>> the gtm and then issue CREATE NODE before you create anothing else. >>>> >>>> >>>> You can visit >>>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>>> failover, start and stop the cluster. init_all function will be helpful >>>> for you. This shows what you should do to initialize your cluster. The >>>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>>> Postgres-XC feature. >>>> >>>> Regards; >>>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Try New Relic Now & We'll Send You this Cool Shirt >>> New Relic is the only SaaS-based application performance monitoring >>> service >>> that delivers powerful full stack analytics. Optimize and monitor your >>> browser, app, & servers with just a few lines of code. Try New Relic >>> and get this awesome Nerd Life shirt! >>> http://p.sf.net/sfu/newrelic_d2d_may >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >>> >> >> >> -- >> StormDB - http://www.stormdb.com >> The Database Cloud >> > > > > ------------------------------------------------------------------------------ > Try New Relic Now & We'll Send You this Cool Shirt > New Relic is the only SaaS-based application performance monitoring service > that delivers powerful full stack analytics. Optimize and monitor your > browser, app, & servers with just a few lines of code. Try New Relic > and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > -- Andrei Martsinchyk StormDB - http://www.stormdb.com The Database Cloud |
|
From: Koichi S. <koi...@gm...> - 2013-05-23 02:36:55
|
The following is a typical additional postgresql.conf file lines for coordinators. They're generated by pgxc_ctl utility and works well. Could you compare this with your ones? Especially, I'm interested how you configured pooler_port. Regards; ----8<----8<---- #=========================================== # Added at initialization. 20130523_11:33:33 port = 20004 pooler_port = 20010 gtm_host = 'node13' gtm_port = 20001 # End of Additon ---->8------->8------ ---------- Koichi Suzuki 2013/5/22 Nikhil Sontakke <ni...@st...> > Hi Donald, > > This looks like a user permissions related internal bug in PGXC. > > Regards, > Nikhils > > > > On Thu, May 23, 2013 at 3:13 AM, Donald Keyass <don...@gm... > > wrote: > >> The testdb and user were the very last thing I created after I created >> the node. >> >> >> On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki <koi...@gm... >> > wrote: >> >>> Did you issue CREATE NODE before you create databases/roles? If you >>> CREATE NODE after you create databases/users, they are not visible to other >>> node and may fall into errors. When you create each node by initdb, you >>> also create gtm's using initgtm, connect all the datanodes/coordinators to >>> the gtm and then issue CREATE NODE before you create anothing else. >>> >>> >>> You can visit >>> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >>> failover, start and stop the cluster. init_all function will be helpful >>> for you. This shows what you should do to initialize your cluster. The >>> C-version of pgxc_ctl will be included in coming 1.1 release with new >>> Postgres-XC feature. >>> >>> Regards; >>> >> >> >> >> ------------------------------------------------------------------------------ >> Try New Relic Now & We'll Send You this Cool Shirt >> New Relic is the only SaaS-based application performance monitoring >> service >> that delivers powerful full stack analytics. Optimize and monitor your >> browser, app, & servers with just a few lines of code. Try New Relic >> and get this awesome Nerd Life shirt! >> http://p.sf.net/sfu/newrelic_d2d_may >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >> >> > > > -- > StormDB - http://www.stormdb.com > The Database Cloud > |
|
From: Nikhil S. <ni...@st...> - 2013-05-22 22:51:23
|
Hi Donald, This looks like a user permissions related internal bug in PGXC. Regards, Nikhils On Thu, May 23, 2013 at 3:13 AM, Donald Keyass <don...@gm...>wrote: > The testdb and user were the very last thing I created after I created the > node. > > > On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki <koi...@gm...>wrote: > >> Did you issue CREATE NODE before you create databases/roles? If you >> CREATE NODE after you create databases/users, they are not visible to other >> node and may fall into errors. When you create each node by initdb, you >> also create gtm's using initgtm, connect all the datanodes/coordinators to >> the gtm and then issue CREATE NODE before you create anothing else. >> >> >> You can visit >> https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctlto look into pgxc_ctl shell script, which takes care of configuration, >> failover, start and stop the cluster. init_all function will be helpful >> for you. This shows what you should do to initialize your cluster. The >> C-version of pgxc_ctl will be included in coming 1.1 release with new >> Postgres-XC feature. >> >> Regards; >> > > > > ------------------------------------------------------------------------------ > Try New Relic Now & We'll Send You this Cool Shirt > New Relic is the only SaaS-based application performance monitoring service > that delivers powerful full stack analytics. Optimize and monitor your > browser, app, & servers with just a few lines of code. Try New Relic > and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > -- StormDB - http://www.stormdb.com The Database Cloud |
|
From: Donald K. <don...@gm...> - 2013-05-22 21:43:20
|
The testdb and user were the very last thing I created after I created the node. On Wed, May 22, 2013 at 4:54 PM, Koichi Suzuki <koi...@gm...>wrote: > Did you issue CREATE NODE before you create databases/roles? If you > CREATE NODE after you create databases/users, they are not visible to other > node and may fall into errors. When you create each node by initdb, you > also create gtm's using initgtm, connect all the datanodes/coordinators to > the gtm and then issue CREATE NODE before you create anothing else. > > > You can visit > https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctl to > look into pgxc_ctl shell script, which takes care of configuration, > failover, start and stop the cluster. init_all function will be helpful > for you. This shows what you should do to initialize your cluster. The > C-version of pgxc_ctl will be included in coming 1.1 release with new > Postgres-XC feature. > > Regards; > |
|
From: Koichi S. <koi...@gm...> - 2013-05-22 20:54:51
|
Did you issue CREATE NODE before you create databases/roles? If you CREATE NODE after you create databases/users, they are not visible to other node and may fall into errors. When you create each node by initdb, you also create gtm's using initgtm, connect all the datanodes/coordinators to the gtm and then issue CREATE NODE before you create anothing else. You can visit https://github.com/koichi-szk/PGXC-Tools/blob/master/pgxc_ctl/pgxc_ctl to look into pgxc_ctl shell script, which takes care of configuration, failover, start and stop the cluster. init_all function will be helpful for you. This shows what you should do to initialize your cluster. The C-version of pgxc_ctl will be included in coming 1.1 release with new Postgres-XC feature. Regards; |
|
From: Donald K. <don...@gm...> - 2013-05-22 19:38:28
|
I'm running pg-xc 1.0.3 on debian linux (squeeze).
I have 3 boxes. 1 for GTM and 2 for datanodes and coordinators.
I have all three up and running. 1 GTM, 2 coordinators and 2 datanodes.
I log in as system user postgres. I can create databases, create users,
select, etc...
So i created a database called testdb. In the testdb, I created a table
called test_tab.
I can query it as postgres user thru both coordinators.
WHen I create a postgresql user:
CREATE ROLE testuser WITH LOGIN PASSWORD 'testpass';
GRANT ALL PRIVILEGES ON DATABASE testdb to testuser;
GRANT CONNECT ON DATABASE testdb TO testuser;
GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA public TO testuser;
I can see the table, but when i select from it, i get:
ERROR: Failed to get pooled connections
I did this on each of the nodes:
/usr/local/pgsql/bin/psql -c "CREATE NODE datanode0 WITH (TYPE =
'datanode', PORT = 5433, HOST = '100.100.10.170')" postgres
/usr/local/pgsql/bin/psql -c "CREATE NODE datanode1 WITH (TYPE =
'datanode', PORT = 5433, HOST = '100.100.10.171')" postgres
/usr/local/pgsql/bin/psql -c "CREATE NODE coord0 WITH (TYPE =
'coordinator', PORT = 5432, HOST = '100.100.10.170')" postgres
/usr/local/pgsql/bin/psql -c "CREATE NODE coord1 WITH (TYPE =
'coordinator', PORT = 5432, HOST = '100.100.10.171')" postgres
and
/usr/local/pgsql/bin/psql -c "SELECT pgxc_pool_reload()" postgres
where i got result of true.
I even did:
postgres=# select pgxc_pool_check();
pgxc_pool_check
-----------------
t
(1 row)
and both are ok.
when i do and lsof -i -n -P
postgres 32131 postgres 3u IPv4 7245254 0t0 TCP *:5433
(LISTEN)
postgres 32131 postgres 4u IPv6 7245255 0t0 TCP *:5433
(LISTEN)
postgres 32141 postgres 3u IPv4 7245283 0t0 TCP *:5432
(LISTEN)
postgres 32141 postgres 4u IPv6 7245284 0t0 TCP *:5432
(LISTEN)
on both servers (data/coord)
So I know they're listening. And two, i was able to create a database and
table and insert and select from it as system user postgres.
But when I create a postgresql user. It doesn't work.
My Coord and DataNode pg_hba.conf is:
# "local" is for Unix domain socket connections only
local all postgres trust
# IPv4 local connections:
host all postgres 100.100.10.170/24 trust
host all postgres 100.100.10.171/24 trust
host all postgres 100.100.10.178/24 trust
#GTM
host all all 127.0.0.1/32 md5
host all all 100.100.10.0/24 md5
# IPv6 local connections:
host all all ::1/128 md5
and in my pgxc_node:
node_name | node_type | node_port | node_host | nodeis_primary |
nodeis_preferred | node_id
-----------+-----------+-----------+-------------+----------------+------------------+-------------
datanode0 | D | 5433 | 100.100.10.170 | f | f
| -96290283
datanode1 | D | 5433 | 100.100.10.171 | f | f
| 888802358
coord1 | C | 5432 | 100.100.10.171 | f | f
| 1885696643
coord0 | C | 5432 | 100.100.10.170 | f | f
| -1944967855
for both servers.
I have them on both coordinator and datanode databases.
I'm @ a loss as to why a system user w/ admin privs can query but a
postgres user w/ priviledges to query gets a :
ERROR: Failed to get pooled connections
I'm figuring it's a config issue somewhere, but i'm not sure as to where.
Can some one point me in the right direction?
Thank you in advance,
Don.
|
|
From: 鈴木 幸市 <ko...@in...> - 2013-05-13 09:51:02
|
I think all the core members and active developers know about this so this will be maintained in the future too. Committers should be careful about it though. Regards; --- Koichi Suzuki On 2013/05/13, at 18:28, Tomonari Katsumata <kat...@po...> wrote: > Hi, Suzuki-san > > According to my test, 1.0.2 works fine, so > I don't have any problem now. > And 1.1dev works fine too. > > I hope nobody will change this feature for future relese. > > thanks, > ----------------- > NTT Software Corporation > Tomonari Katsumata > > >> Katsumata-san; >> >> Unfortunately it was not included in 1.0.2 to save development > resources. I'd like to know how serious it is for 1.0, not for 1.1. > Depending upon severity, I may be able to re-prioritize my resource. >> >> Regards; >> --- >> Koichi Suzuki >> >> >> >> On 2013/05/13, at 14:39, Tomonari Katsumata > <kat...@po...> wrote: >> >>> Hi, >>> >>> I'm thinking about how to recover the system after GTM crash. >>> >>> After GTM stopped, I just start it again. >>> It works fine with Postgres-XC 1.0.2, and I could do "CREATE TABLE", >>> "INSERT" some data and "SELECT" them. >>> >>> But with Postgres-XC 1.0.1, It doesn't work same. >>> I have to make GTM_proxy reconnect to GTM manually(gtm_ctl reconnect). >>> In some cases, I have to restart GTM_proxy. >>> >>> I want to know 2 things about this. >>> 1. I think this new feature(*) is included in Postgres-XC 1.0.2, right? >>> (*) reconnect doesn't be required with restarting GTM. >>> 2. May I think this feature will never change for future releases? >>> (I've confirmed the feature is alive in Postgres-XC 1.1dev !) >>> >>> ============= >>> Environment >>> ============= >>> 2 servers are there. >>> components are destributed like bellow. >>> >>> server1 - GTM, GTM_proxy1, Coordinator1, Datanode1 >>> server2 - GTM_proxy2, Coordinator2, Datanode2 >>> >>> ========== >>> reproduce >>> ========== >>> 1. stop GTM manually (kill -9 <PID of GTM>) >>> 2. start GTM again manually. >>> >>> >>> regards, >>> ----------------- >>> NTT Software Corporation >>> Tomonari Katsumata >>> >>> >>> >>> > ------------------------------------------------------------------------------ >>> Learn Graph Databases - Download FREE O'Reilly Book >>> "Graph Databases" is the definitive new guide to graph databases and >>> their applications. This 200-page book is written by three acclaimed >>> leaders in the field. The early access version is available now. >>> Download your free book today! http://p.sf.net/sfu/neotech_d2d_may >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >> >> > > > > > > ------------------------------------------------------------------------------ > Learn Graph Databases - Download FREE O'Reilly Book > "Graph Databases" is the definitive new guide to graph databases and > their applications. This 200-page book is written by three acclaimed > leaders in the field. The early access version is available now. > Download your free book today! http://p.sf.net/sfu/neotech_d2d_may > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > |
|
From: Tomonari K. <kat...@po...> - 2013-05-13 09:32:23
|
Hi, Suzuki-san According to my test, 1.0.2 works fine, so I don't have any problem now. And 1.1dev works fine too. I hope nobody will change this feature for future relese. thanks, ----------------- NTT Software Corporation Tomonari Katsumata > Katsumata-san; > > Unfortunately it was not included in 1.0.2 to save development resources. I'd like to know how serious it is for 1.0, not for 1.1. Depending upon severity, I may be able to re-prioritize my resource. > > Regards; > --- > Koichi Suzuki > > > > On 2013/05/13, at 14:39, Tomonari Katsumata <kat...@po...> wrote: > >> Hi, >> >> I'm thinking about how to recover the system after GTM crash. >> >> After GTM stopped, I just start it again. >> It works fine with Postgres-XC 1.0.2, and I could do "CREATE TABLE", >> "INSERT" some data and "SELECT" them. >> >> But with Postgres-XC 1.0.1, It doesn't work same. >> I have to make GTM_proxy reconnect to GTM manually(gtm_ctl reconnect). >> In some cases, I have to restart GTM_proxy. >> >> I want to know 2 things about this. >> 1. I think this new feature(*) is included in Postgres-XC 1.0.2, right? >> (*) reconnect doesn't be required with restarting GTM. >> 2. May I think this feature will never change for future releases? >> (I've confirmed the feature is alive in Postgres-XC 1.1dev !) >> >> ============= >> Environment >> ============= >> 2 servers are there. >> components are destributed like bellow. >> >> server1 - GTM, GTM_proxy1, Coordinator1, Datanode1 >> server2 - GTM_proxy2, Coordinator2, Datanode2 >> >> ========== >> reproduce >> ========== >> 1. stop GTM manually (kill -9 <PID of GTM>) >> 2. start GTM again manually. >> >> >> regards, >> ----------------- >> NTT Software Corporation >> Tomonari Katsumata >> >> >> >> ------------------------------------------------------------------------------ >> Learn Graph Databases - Download FREE O'Reilly Book >> "Graph Databases" is the definitive new guide to graph databases and >> their applications. This 200-page book is written by three acclaimed >> leaders in the field. The early access version is available now. >> Download your free book today! http://p.sf.net/sfu/neotech_d2d_may >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >> > > |
|
From: 鈴木 幸市 <ko...@in...> - 2013-05-13 07:03:47
|
Katsumata-san; Unfortunately it was not included in 1.0.2 to save development resources. I'd like to know how serious it is for 1.0, not for 1.1. Depending upon severity, I may be able to re-prioritize my resource. Regards; --- Koichi Suzuki On 2013/05/13, at 14:39, Tomonari Katsumata <kat...@po...> wrote: > Hi, > > I'm thinking about how to recover the system after GTM crash. > > After GTM stopped, I just start it again. > It works fine with Postgres-XC 1.0.2, and I could do "CREATE TABLE", > "INSERT" some data and "SELECT" them. > > But with Postgres-XC 1.0.1, It doesn't work same. > I have to make GTM_proxy reconnect to GTM manually(gtm_ctl reconnect). > In some cases, I have to restart GTM_proxy. > > I want to know 2 things about this. > 1. I think this new feature(*) is included in Postgres-XC 1.0.2, right? > (*) reconnect doesn't be required with restarting GTM. > 2. May I think this feature will never change for future releases? > (I've confirmed the feature is alive in Postgres-XC 1.1dev !) > > ============= > Environment > ============= > 2 servers are there. > components are destributed like bellow. > > server1 - GTM, GTM_proxy1, Coordinator1, Datanode1 > server2 - GTM_proxy2, Coordinator2, Datanode2 > > ========== > reproduce > ========== > 1. stop GTM manually (kill -9 <PID of GTM>) > 2. start GTM again manually. > > > regards, > ----------------- > NTT Software Corporation > Tomonari Katsumata > > > > ------------------------------------------------------------------------------ > Learn Graph Databases - Download FREE O'Reilly Book > "Graph Databases" is the definitive new guide to graph databases and > their applications. This 200-page book is written by three acclaimed > leaders in the field. The early access version is available now. > Download your free book today! http://p.sf.net/sfu/neotech_d2d_may > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > |
|
From: Tomonari K. <kat...@po...> - 2013-05-13 05:43:22
|
Hi, I'm thinking about how to recover the system after GTM crash. After GTM stopped, I just start it again. It works fine with Postgres-XC 1.0.2, and I could do "CREATE TABLE", "INSERT" some data and "SELECT" them. But with Postgres-XC 1.0.1, It doesn't work same. I have to make GTM_proxy reconnect to GTM manually(gtm_ctl reconnect). In some cases, I have to restart GTM_proxy. I want to know 2 things about this. 1. I think this new feature(*) is included in Postgres-XC 1.0.2, right? (*) reconnect doesn't be required with restarting GTM. 2. May I think this feature will never change for future releases? (I've confirmed the feature is alive in Postgres-XC 1.1dev !) ============= Environment ============= 2 servers are there. components are destributed like bellow. server1 - GTM, GTM_proxy1, Coordinator1, Datanode1 server2 - GTM_proxy2, Coordinator2, Datanode2 ========== reproduce ========== 1. stop GTM manually (kill -9 <PID of GTM>) 2. start GTM again manually. regards, ----------------- NTT Software Corporation Tomonari Katsumata |
|
From: Adam D. <ada...@sm...> - 2013-05-12 16:12:20
|
# ---------------------- # GTM configuration file # ---------------------- # # This file must be placed on gtm working directory # specified by -D command line option of gtm or gtm_ctl. The # configuration file name must be "gtm.conf" # # # This file consists of lines of the form # # name = value # # (The "=" is optional.) Whitespace may be used. Comments are # introduced with "#" anywhere on a line. The complete list of # parameter names and allowed values can be found in the # Postgres-XC documentation. # # The commented-out settings shown in this file represent the default # values. # # Re-commenting a setting is NOT sufficient to revert it to the default # value. # # You need to restart the server. #------------------------------------------------------------------------------ # GENERAL PARAMETERS #------------------------------------------------------------------------------ nodename = 'GTM1' # Specifies the node name. # (changes requires restart) listen_addresses = '*' # Listen addresses of this GTM. # (changes requires restart) port = 20001 # Port number of this GTM. # (changes requires restart) #startup = ACT # Start mode. ACT/STANDBY. #------------------------------------------------------------------------------ # GTM STANDBY PARAMETERS #------------------------------------------------------------------------------ #Those parameters are effective when GTM is activated as a standby server #active_host = '' # Listen address of active GTM. # (changes requires restart) #active_port = # Port number of active GTM. # (changes requires restart) #--------------------------------------- # OTHER OPTIONS #--------------------------------------- #keepalives_idle = 0 # Keepalives_idle parameter. #keepalives_interval = 0 # Keepalives_interval parameter. #keepalives_count = 0 # Keepalives_count internal parameter. #log_file = 'gtm.log' # Log file name #log_min_messages = DEBUG5 # log_min_messages. Default WARNING. # Valid value: DEBUG, DEBUG5, DEBUG4, DEBUG3, # DEBUG2, DEBUG1, INFO, NOTICE, WARNING, # ERROR, LOG, FATAL, PANIC #synchronous_backup = off # If backup to standby is synchronous |
|
From: 鈴木 幸市 <ko...@in...> - 2013-05-09 05:53:02
|
Thanx. --- Koichi Suzuki On 2013/05/08, at 14:46, "Masaki HISADA" <his...@la...> wrote: > Hi, Koichi-san, > >> Could you open bug tickets in sourceforge and assign them to me? > > ID: 3612883 > > Rgds, > > Hisada > >> -----Original Message----- >> From: 鈴木 幸市 [mailto:ko...@in...] >> Sent: Wednesday, May 01, 2013 9:51 AM >> To: Masaki HISADA >> Cc: pos...@li... >> Subject: Re: [Postgres-xc-general] GTM standby behaivor when cannot > connect >> with GTM active >> >> I have also observed connection error message in gtm slave although > there're >> no network failure and gtm master is running. We should look into it. >> >> Also failure to stop gtm standby could be a bug. >> >> Could you open bug tickets in sourceforge and assign them to me? >> --- >> Koichi Suzuki >> >> >> >> On 2013/04/30, at 16:04, Masaki HISADA <his...@la...> > wrote: >> >>> Hi Kohichi-san, >>> >>> >>>> At present, gtm_standby does not use tcp_keepalives. Gtm_standby > just >>> wait >>>> for GTM to send backups and does not detect network fault until the >>>> kernel timeout (usually 60min or so). >>> >>> From gtm.log, it seems that gtm_standby detecting connection failure >>> with GTM about 1 minute after starting up gtm_standby. >>> >>>>> $ less gtm/data/gtm.log >>>>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the >>>>> GTM active on 192.168.2.1:19000... >>>>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 >>>>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to >>>>> establish a connection to active-GTM. >>>>> LOCATION: main, main.c:553 >>> >>>> Because it continues to wait for incoming backup commands, status >>>> will >>> return >>>> that it's running. If Gtm slave starts during the network fault, it >>> cannot >>>> get initial response from master and it will fail. >>> >>> I have tried stop immediate in this situation but could not stop >>> gtm_standby but cannot stop gtm_standby .... >>> Although gtm_standby is waiting for GTM backup, I guess gtm_standby >>> needs to be shutting down when immediate shutdown. >>> >>>> (4) Stop gtm_stadby >>> >>> $ gtm_ctl -D gtm/data/ -Z gtm start >>> server starting >>> $ gtm_ctl -D gtm/data/ -Z gtm stop -m immediate waiting for server to >>> shut down................ >>> ............................................... failed >>> gtm_ctl: server does not shut down >>> >>> Regards, >>> >>> Mark >>> >>> >>>> Regards; >>>> --- >>>> Koichi Suzuki >>>> >>>> >>>> >>>> On 2013/04/26, at 16:00, Masaki HISADA <his...@la...> >>> wrote: >>>> >>>>> Hi, >>>>> >>>>> I have been testing pgxc behavior when network between gtm and >>>>> gtm_standy has down. >>>>> >>>>> I have started up gtm_standby and linked down bw gtm and gtm_standby >>>>> by using iptables. I expected that gtm_standby would fail with error >>>>> because gtm_standby cannot establish synchronization with gtm active. >>>>> >>>>> Problem : >>>>> 1. gtm_standby started without error response although gtm_standby >>>>> has failed to establish connection with gtm active from gtm.log. Is >>>>> gtm alive or not? >>>>> 2. Assuming gtm alive I have tried immediate stop to gtm_standby but >>> failed. >>>>> 3. Checking at gtm status, it is stopped. >>>>> >>>>> Question >>>>> (1) Am I testing right? >>>>> (2) Is this expected behavior for gtm_standby when it cannot >>>>> establish connection with gtm? >>>>> >>>>> Test procedure as follows; >>>>> >>>>> (1) Start up GTM active at Server A >>>>> $ gtm_ctl -D gtm/data/ -Z gtm status >>>>> >>>>> (2) Network down (simulated with setting iptables) $ iptables -I >>>>> INPUT -i bond1 -j DROP $ iptables -I OUTPUT -o bond1 -j DROP >>>>> >>>>> (3) Start up GTM standby at Server B $ gtm_ctl -D gtm/data/ -Z gtm >>>>> start server starting >>>>> >>>>> (4) Stop gtm_stadby >>>>> waiting for server to shut down................ >>>>> ............................................... failed >>>>> gtm_ctl: server does not shut down >>>>> >>>>> (5) Check status >>>>> $ gtm_ctl -D gtm/data/ -Z gtm status >>>>> gtm_ctl: no server running >>>>> $ ps aux | grep gtm | grep -v grep >>>>> # No output >>>>> $ ll gtm/data/ >>>>> 合計 12 >>>>> -rw------- 1 pgxc pgxc 2217 2月 15 06:15 2013 gtm.conf >>>>> -rw------- 1 pgxc pgxc 295 2月 25 04:39 2013 gtm.log >>>>> -rw------- 1 pgxc pgxc 45 2月 25 04:38 2013 gtm.pid >>>>> >>>>> $ less gtm/data/gtm.log >>>>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the >>>>> GTM active on 192.168.2.1:19000... >>>>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 >>>>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to >>>>> establish a connection to active-GTM. >>>>> LOCATION: main, main.c:553 >>>>> >>>>> pos...@li... >>>>> >>>>> Tested Version : commitID(3f15aef4ed3f8b2014d42bce3ed75650c4db907d) >>>>> >>>>> Rgds, >>>>> >>>>> Mark >>>>> >>>>> >>>>> -------------------------------------------------------------------- >>>>> -- >>>>> -------- Try New Relic Now & We'll Send You this Cool Shirt New >>>>> Relic is the only SaaS-based application performance monitoring >>>>> service that delivers powerful full stack analytics. Optimize and >>>>> monitor your browser, app, & servers with just a few lines of code. >>>>> Try New Relic and get this awesome Nerd Life shirt! >>>>> http://p.sf.net/sfu/newrelic_d2d_apr >>>>> _______________________________________________ >>>>> Postgres-xc-general mailing list >>>>> Pos...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>>>> >>>> >>>> >>>> --------------------------------------------------------------------- >>>> ---- >>>> ----- >>>> Introducing AppDynamics Lite, a free troubleshooting tool for >>>> Java/.NET >>> Get >>>> 100% visibility into your production application - at no cost. >>>> Code-level diagnostics for performance bottlenecks with <2% overhead >>> Download >>>> for free and get started troubleshooting in minutes. >>>> http://p.sf.net/sfu/appdyn_d2d_ap1 >>>> _______________________________________________ >>>> Postgres-xc-general mailing list >>>> Pos...@li... >>>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >>> >> >> >> ------------------------------------------------------------------------- >> ----- >> Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET > Get >> 100% visibility into your production application - at no cost. >> Code-level diagnostics for performance bottlenecks with <2% overhead > Download >> for free and get started troubleshooting in minutes. >> http://p.sf.net/sfu/appdyn_d2d_ap1 >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > |
|
From: Masaki H. <his...@la...> - 2013-05-08 05:45:59
|
Hi, Koichi-san, > Could you open bug tickets in sourceforge and assign them to me? ID: 3612883 Rgds, Hisada > -----Original Message----- > From: 鈴木 幸市 [mailto:ko...@in...] > Sent: Wednesday, May 01, 2013 9:51 AM > To: Masaki HISADA > Cc: pos...@li... > Subject: Re: [Postgres-xc-general] GTM standby behaivor when cannot connect > with GTM active > > I have also observed connection error message in gtm slave although there're > no network failure and gtm master is running. We should look into it. > > Also failure to stop gtm standby could be a bug. > > Could you open bug tickets in sourceforge and assign them to me? > --- > Koichi Suzuki > > > > On 2013/04/30, at 16:04, Masaki HISADA <his...@la...> wrote: > > > Hi Kohichi-san, > > > > > >> At present, gtm_standby does not use tcp_keepalives. Gtm_standby just > > wait > >> for GTM to send backups and does not detect network fault until the > >> kernel timeout (usually 60min or so). > > > > From gtm.log, it seems that gtm_standby detecting connection failure > > with GTM about 1 minute after starting up gtm_standby. > > > >>> $ less gtm/data/gtm.log > >>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the > >>> GTM active on 192.168.2.1:19000... > >>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 > >>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to > >>> establish a connection to active-GTM. > >>> LOCATION: main, main.c:553 > > > >> Because it continues to wait for incoming backup commands, status > >> will > > return > >> that it's running. If Gtm slave starts during the network fault, it > > cannot > >> get initial response from master and it will fail. > > > > I have tried stop immediate in this situation but could not stop > > gtm_standby but cannot stop gtm_standby .... > > Although gtm_standby is waiting for GTM backup, I guess gtm_standby > > needs to be shutting down when immediate shutdown. > > > >> (4) Stop gtm_stadby > > > > $ gtm_ctl -D gtm/data/ -Z gtm start > > server starting > > $ gtm_ctl -D gtm/data/ -Z gtm stop -m immediate waiting for server to > > shut down................ > > ............................................... failed > > gtm_ctl: server does not shut down > > > > Regards, > > > > Mark > > > > > >> Regards; > >> --- > >> Koichi Suzuki > >> > >> > >> > >> On 2013/04/26, at 16:00, Masaki HISADA <his...@la...> > > wrote: > >> > >>> Hi, > >>> > >>> I have been testing pgxc behavior when network between gtm and > >>> gtm_standy has down. > >>> > >>> I have started up gtm_standby and linked down bw gtm and gtm_standby > >>> by using iptables. I expected that gtm_standby would fail with error > >>> because gtm_standby cannot establish synchronization with gtm active. > >>> > >>> Problem : > >>> 1. gtm_standby started without error response although gtm_standby > >>> has failed to establish connection with gtm active from gtm.log. Is > >>> gtm alive or not? > >>> 2. Assuming gtm alive I have tried immediate stop to gtm_standby but > > failed. > >>> 3. Checking at gtm status, it is stopped. > >>> > >>> Question > >>> (1) Am I testing right? > >>> (2) Is this expected behavior for gtm_standby when it cannot > >>> establish connection with gtm? > >>> > >>> Test procedure as follows; > >>> > >>> (1) Start up GTM active at Server A > >>> $ gtm_ctl -D gtm/data/ -Z gtm status > >>> > >>> (2) Network down (simulated with setting iptables) $ iptables -I > >>> INPUT -i bond1 -j DROP $ iptables -I OUTPUT -o bond1 -j DROP > >>> > >>> (3) Start up GTM standby at Server B $ gtm_ctl -D gtm/data/ -Z gtm > >>> start server starting > >>> > >>> (4) Stop gtm_stadby > >>> waiting for server to shut down................ > >>> ............................................... failed > >>> gtm_ctl: server does not shut down > >>> > >>> (5) Check status > >>> $ gtm_ctl -D gtm/data/ -Z gtm status > >>> gtm_ctl: no server running > >>> $ ps aux | grep gtm | grep -v grep > >>> # No output > >>> $ ll gtm/data/ > >>> 合計 12 > >>> -rw------- 1 pgxc pgxc 2217 2月 15 06:15 2013 gtm.conf > >>> -rw------- 1 pgxc pgxc 295 2月 25 04:39 2013 gtm.log > >>> -rw------- 1 pgxc pgxc 45 2月 25 04:38 2013 gtm.pid > >>> > >>> $ less gtm/data/gtm.log > >>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the > >>> GTM active on 192.168.2.1:19000... > >>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 > >>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to > >>> establish a connection to active-GTM. > >>> LOCATION: main, main.c:553 > >>> > >>> pos...@li... > >>> > >>> Tested Version : commitID(3f15aef4ed3f8b2014d42bce3ed75650c4db907d) > >>> > >>> Rgds, > >>> > >>> Mark > >>> > >>> > >>> -------------------------------------------------------------------- > >>> -- > >>> -------- Try New Relic Now & We'll Send You this Cool Shirt New > >>> Relic is the only SaaS-based application performance monitoring > >>> service that delivers powerful full stack analytics. Optimize and > >>> monitor your browser, app, & servers with just a few lines of code. > >>> Try New Relic and get this awesome Nerd Life shirt! > >>> http://p.sf.net/sfu/newrelic_d2d_apr > >>> _______________________________________________ > >>> Postgres-xc-general mailing list > >>> Pos...@li... > >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > >>> > >> > >> > >> --------------------------------------------------------------------- > >> ---- > >> ----- > >> Introducing AppDynamics Lite, a free troubleshooting tool for > >> Java/.NET > > Get > >> 100% visibility into your production application - at no cost. > >> Code-level diagnostics for performance bottlenecks with <2% overhead > > Download > >> for free and get started troubleshooting in minutes. > >> http://p.sf.net/sfu/appdyn_d2d_ap1 > >> _______________________________________________ > >> Postgres-xc-general mailing list > >> Pos...@li... > >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > > > > > > ------------------------------------------------------------------------- > ----- > Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET Get > 100% visibility into your production application - at no cost. > Code-level diagnostics for performance bottlenecks with <2% overhead Download > for free and get started troubleshooting in minutes. > http://p.sf.net/sfu/appdyn_d2d_ap1 > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general |
|
From: 鈴木 幸市 <ko...@in...> - 2013-05-02 05:28:59
|
Thanks Tim for the info. I visited the site and took a look at the patch they provide for postgres-xc core. I'm afraid the patch may not for general use because they ignore some kind of errors. Original code is to abort the current transaction but modified patch requires to ignore it. It's helpful if StormDB people or other people review and test to show this patch works generally and doesn't bring any issues. Maybe PostGis installation issues Postgres-XC SQL statements and they complain that there're no datanode. The patch looks like a workaround for them. The installation and configuration procedure assumes that all the components run on the same server, using identical binary. If you have more than one server, installation may need something more. If it is agreed that the patch is for general use, it can be included in XC core. Regards; --- Koichi Suzuki On 2013/05/01, at 12:17, Tim Keitt <tk...@ut...> wrote: > Here's the link to postgis-postgres-xc integration: > > http://goo.gl/JHHjH > > THK > > -- > http://www.keittlab.org/ > ------------------------------------------------------------------------------ > Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET > Get 100% visibility into your production application - at no cost. > Code-level diagnostics for performance bottlenecks with <2% overhead > Download for free and get started troubleshooting in minutes. > http://p.sf.net/sfu/appdyn_d2d_ap1_______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general |
|
From: Tim K. <tk...@ut...> - 2013-05-01 03:17:57
|
Here's the link to postgis-postgres-xc integration: http://goo.gl/JHHjH THK -- http://www.keittlab.org/ |
|
From: 鈴木 幸市 <ko...@in...> - 2013-05-01 00:51:36
|
I have also observed connection error message in gtm slave although there're no network failure and gtm master is running. We should look into it. Also failure to stop gtm standby could be a bug. Could you open bug tickets in sourceforge and assign them to me? --- Koichi Suzuki On 2013/04/30, at 16:04, Masaki HISADA <his...@la...> wrote: > Hi Kohichi-san, > > >> At present, gtm_standby does not use tcp_keepalives. Gtm_standby just > wait >> for GTM to send backups and does not detect network fault until the kernel >> timeout (usually 60min or so). > > From gtm.log, it seems that gtm_standby detecting connection failure with > GTM about 1 minute after starting up gtm_standby. > >>> $ less gtm/data/gtm.log >>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the >>> GTM active on 192.168.2.1:19000... >>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 >>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to >>> establish a connection to active-GTM. >>> LOCATION: main, main.c:553 > >> Because it continues to wait for incoming backup commands, status will > return >> that it's running. If Gtm slave starts during the network fault, it > cannot >> get initial response from master and it will fail. > > I have tried stop immediate in this situation but could not stop gtm_standby > but cannot stop gtm_standby .... > Although gtm_standby is waiting for GTM backup, I guess gtm_standby needs to > be shutting down when immediate shutdown. > >> (4) Stop gtm_stadby > > $ gtm_ctl -D gtm/data/ -Z gtm start > server starting > $ gtm_ctl -D gtm/data/ -Z gtm stop -m immediate > waiting for server to shut down................ > ............................................... failed > gtm_ctl: server does not shut down > > Regards, > > Mark > > >> Regards; >> --- >> Koichi Suzuki >> >> >> >> On 2013/04/26, at 16:00, Masaki HISADA <his...@la...> > wrote: >> >>> Hi, >>> >>> I have been testing pgxc behavior when network between gtm and >>> gtm_standy has down. >>> >>> I have started up gtm_standby and linked down bw gtm and gtm_standby >>> by using iptables. I expected that gtm_standby would fail with error >>> because gtm_standby cannot establish synchronization with gtm active. >>> >>> Problem : >>> 1. gtm_standby started without error response although gtm_standby has >>> failed to establish connection with gtm active from gtm.log. Is gtm >>> alive or not? >>> 2. Assuming gtm alive I have tried immediate stop to gtm_standby but > failed. >>> 3. Checking at gtm status, it is stopped. >>> >>> Question >>> (1) Am I testing right? >>> (2) Is this expected behavior for gtm_standby when it cannot establish >>> connection with gtm? >>> >>> Test procedure as follows; >>> >>> (1) Start up GTM active at Server A >>> $ gtm_ctl -D gtm/data/ -Z gtm status >>> >>> (2) Network down (simulated with setting iptables) $ iptables -I INPUT >>> -i bond1 -j DROP $ iptables -I OUTPUT -o bond1 -j DROP >>> >>> (3) Start up GTM standby at Server B >>> $ gtm_ctl -D gtm/data/ -Z gtm start >>> server starting >>> >>> (4) Stop gtm_stadby >>> waiting for server to shut down................ >>> ............................................... failed >>> gtm_ctl: server does not shut down >>> >>> (5) Check status >>> $ gtm_ctl -D gtm/data/ -Z gtm status >>> gtm_ctl: no server running >>> $ ps aux | grep gtm | grep -v grep >>> # No output >>> $ ll gtm/data/ >>> 合計 12 >>> -rw------- 1 pgxc pgxc 2217 2月 15 06:15 2013 gtm.conf >>> -rw------- 1 pgxc pgxc 295 2月 25 04:39 2013 gtm.log >>> -rw------- 1 pgxc pgxc 45 2月 25 04:38 2013 gtm.pid >>> >>> $ less gtm/data/gtm.log >>> 1:140357722076928:2013-02-25 04:38:23.850 JST -LOG: Connecting the >>> GTM active on 192.168.2.1:19000... >>> LOCATION: gtm_standby_connectToActiveGTM, gtm_standby.c:511 >>> 1:140357722076928:2013-02-25 04:39:26.856 JST -FATAL: Failed to >>> establish a connection to active-GTM. >>> LOCATION: main, main.c:553 >>> >>> pos...@li... >>> >>> Tested Version : commitID(3f15aef4ed3f8b2014d42bce3ed75650c4db907d) >>> >>> Rgds, >>> >>> Mark >>> >>> >>> ---------------------------------------------------------------------- >>> -------- Try New Relic Now & We'll Send You this Cool Shirt New Relic >>> is the only SaaS-based application performance monitoring service that >>> delivers powerful full stack analytics. Optimize and monitor your >>> browser, app, & servers with just a few lines of code. Try New Relic >>> and get this awesome Nerd Life shirt! >>> http://p.sf.net/sfu/newrelic_d2d_apr >>> _______________________________________________ >>> Postgres-xc-general mailing list >>> Pos...@li... >>> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >>> >> >> >> ------------------------------------------------------------------------- >> ----- >> Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET > Get >> 100% visibility into your production application - at no cost. >> Code-level diagnostics for performance bottlenecks with <2% overhead > Download >> for free and get started troubleshooting in minutes. >> http://p.sf.net/sfu/appdyn_d2d_ap1 >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > |
1 message has been excluded from this view by a project administrator.
