Thread: Postmaster re-start problem
I was just about finished transferring a database from a PostreSQL 6.5.2
installation on server A to a PostgreSQL 7.0.3 installation on server B
when I ran into a bit of trouble. Both installations are on RedHat 6.2
boxes. Here's what I did...
* On server A: ran the pg_dump command and dumped the database I
wanted to an output file.
* Transferred the output file from server A to server B.
* On server B: ran the following command: "psql -d timeclock -f
db.out -U user_manager -W". Everything completed successfully. I
opened and interactive session as user_manager and made sure
everything was there.
* On server B: edited the pg_hba.conf file to allow connectivity
from a single external host by adding the entry "host timeclock
192.168.9.30 255.255.255.255 crypt"
* On server B: ran the "/etc/rc.d/init.d/postgres restart" command.
This is were the problem lies - the post master won't restart.
I took a look at the output of the "ps ax" command and I found the
following entry:
17756 ? R 29:09 /usr/bin/postgres localhost
user_manager timeclock id
Am I correct in assuming that this is prohibiting a restart? If so, is
it OK to kill this process off? If I can't kill it off, what are my
options?
Thanks.
--
John Burski
I.T. Manager and Systems Administration
911 Emergency Products, Inc.
25 Sixth Avenue North
Saint Cloud, MN 56303
John.Burski@911ep.com
800-863-6911, extension 221
FAX: 800-863-2991
www.911ep.com
Thanks for the response. I had some trouble posting the question to the group - the list didn't think I was a member. In any case, I killed the process off without any adverse affects. I gathered from your response that the "pg_hba.conf" file isn't read at postmaster startup exclusively, but, rather, whenever authentication is required. I, obviously, wasn't aware of that - thanks for the education. :) Ciao! Peter Eisentraut wrote: > John Burski writes: > > > I was just about finished transferring a database from a PostreSQL 6.5.2 > > installation on server A to a PostgreSQL 7.0.3 installation on server B > > Try 7.1.3 instead. > > > when I ran into a bit of trouble. Both installations are on RedHat 6.2 > > boxes. Here's what I did... > > > > * On server A: ran the pg_dump command and dumped the database I > > wanted to an output file. > > * Transferred the output file from server A to server B. > > * On server B: ran the following command: "psql -d timeclock -f > > db.out -U user_manager -W". Everything completed successfully. I > > opened and interactive session as user_manager and made sure > > everything was there. > > * On server B: edited the pg_hba.conf file to allow connectivity > > from a single external host by adding the entry "host timeclock > > 192.168.9.30 255.255.255.255 crypt" > > Up to here would have been sufficient. > > > * On server B: ran the "/etc/rc.d/init.d/postgres restart" command. > > This is were the problem lies - the post master won't restart. > > This is a complete blind guess without knowing what this > /etc/rc.d/init.d/postgres file contains. > > > I took a look at the output of the "ps ax" command and I found the > > following entry: > > > > 17756 ? R 29:09 /usr/bin/postgres localhost > > user_manager timeclock id > > > > Am I correct in assuming that this is prohibiting a restart? > > Possibly, see above. > > > If so, is it OK to kill this process off? > > If you don't think anyone is or should be using the database, yes. > > > If I can't kill it off, what are my options? > > Turn off your computer? ;-) > > -- > Peter Eisentraut peter_e@gmx.net http://funkturm.homeip.net/~peter -- John Burski I.T. Manager and Systems Administration 911 Emergency Products, Inc. 25 Sixth Avenue North Saint Cloud, MN 56303 John.Burski@911ep.com 800-863-6911, extension 221 FAX: 800-863-2991 www.911ep.com
> Thanks for the response. > > I had some trouble posting the question to the group - the list didn't think > I was a member. In any case, I killed the process off without any adverse > affects. > > I gathered from your response that the "pg_hba.conf" file isn't read at > postmaster startup exclusively, but, rather, whenever authentication is > required. I, obviously, wasn't aware of that - thanks for the education. :) This will change in 7.2 and pg_hba.conf will be read only on start or SIGHUP. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
John Burski writes: > I was just about finished transferring a database from a PostreSQL 6.5.2 > installation on server A to a PostgreSQL 7.0.3 installation on server B Try 7.1.3 instead. > when I ran into a bit of trouble. Both installations are on RedHat 6.2 > boxes. Here's what I did... > > * On server A: ran the pg_dump command and dumped the database I > wanted to an output file. > * Transferred the output file from server A to server B. > * On server B: ran the following command: "psql -d timeclock -f > db.out -U user_manager -W". Everything completed successfully. I > opened and interactive session as user_manager and made sure > everything was there. > * On server B: edited the pg_hba.conf file to allow connectivity > from a single external host by adding the entry "host timeclock > 192.168.9.30 255.255.255.255 crypt" Up to here would have been sufficient. > * On server B: ran the "/etc/rc.d/init.d/postgres restart" command. > This is were the problem lies - the post master won't restart. This is a complete blind guess without knowing what this /etc/rc.d/init.d/postgres file contains. > I took a look at the output of the "ps ax" command and I found the > following entry: > > 17756 ? R 29:09 /usr/bin/postgres localhost > user_manager timeclock id > > Am I correct in assuming that this is prohibiting a restart? Possibly, see above. > If so, is it OK to kill this process off? If you don't think anyone is or should be using the database, yes. > If I can't kill it off, what are my options? Turn off your computer? ;-) -- Peter Eisentraut peter_e@gmx.net http://funkturm.homeip.net/~peter