Обсуждение: Server Won't Start On a Copy of Its Own Data Folder
Server Won't Start On a Copy of Its Own Data Folder
1. Installed the latest 8.0.3 binary on a Windows XP Pro system as a service to be started on bootup.
2. Used psql just to verify that I can connect.
3. Used Windows Control Panel.Administrative Tools.Services to stop the server cleanly.
4. Made a copy of the PostgreSQL/8.0/data folder
5. Renamed the installed folder as PostgreSQL/8.0/saved-data
6. Renamed the copy of data as PostrgeSQL/8.0/data
7. Used Windows Control Panel.Administrative Tools.Services to start the server
8. The server fails to come up
Windows comes back and tells me that the service started and stopped. The Windows
EventLog show that the postgres server was sent a start control, entered the running state and
then entered the stopped state. There was nothing new in the pg_log folder to provide
any clue.
I renamed the saved-data folder back to data and used the Services tool to start the server
with success but further attemps to start it with a copy of the data folder all fail in the same way.
Based on what I've gleaned on reading some of the past mail archives, I thought that this should
work. Have I misunderstood something?
I went through this exercise in trying to solve a problem I posted about on Monday
([ADMIN] Trouble Getting Database Server Started) trying to recover a database on
a computer that may have powered down when the server was running. I don't know
what happened but I can't start the server any more on that computer. My goal is to
install the data folder from that computer on a new install of postgres on a second
computer.
Thanks for any help anyone can offer,
Dirk Kalp
> Server Won't Start On a Copy of Its Own Data Folder > > 1. Installed the latest 8.0.3 binary on a Windows XP Pro > system as a service to be started on bootup. > 2. Used psql just to verify that I can connect. > 3. Used Windows Control Panel.Administrative Tools.Services > to stop the server cleanly. > 4. Made a copy of the PostgreSQL/8.0/data folder > 5. Renamed the installed folder as PostgreSQL/8.0/saved-data > 6. Renamed the copy of data as PostrgeSQL/8.0/data > 7. Used Windows Control Panel.Administrative Tools.Services > to start the server > 8. The server fails to come up > > Windows comes back and tells me that the service started and > stopped. The Windows > EventLog show that the postgres server was sent a start > control, entered the running state and > then entered the stopped state. There was nothing new in the > pg_log folder to provide > any clue. This is very weird - it should normally log *something*. There is nothing from postgresql in the eventlog? > Based on what I've gleaned on reading some of the past mail > archives, I thought that this should > work. Have I misunderstood something? It should,m as long as they are on the same version. And it should absolutely log an error. Can you try starting the database manually using pg_ctl (not as a service) and see if you get more errors then? Remember to start it from a runas:ed shell, so you run it as the same user. //Magnus
--- Magnus Hagander <mha@sollentuna.net> wrote: > > Server Won't Start On a Copy of Its Own Data > Folder > > > > 1. Installed the latest 8.0.3 binary on a Windows > XP Pro > > system as a service to be started on bootup. > > 2. Used psql just to verify that I can connect. > > 3. Used Windows Control Panel.Administrative > Tools.Services > > to stop the server cleanly. > > 4. Made a copy of the PostgreSQL/8.0/data folder > > 5. Renamed the installed folder as > PostgreSQL/8.0/saved-data > > 6. Renamed the copy of data as PostrgeSQL/8.0/data > > 7. Used Windows Control Panel.Administrative > Tools.Services > > to start the server > > 8. The server fails to come up > > > > Windows comes back and tells me that the service > started and > > stopped. The Windows > > EventLog show that the postgres server was sent a > start > > control, entered the running state and > > then entered the stopped state. There was nothing > new in the > > pg_log folder to provide > > any clue. > > This is very weird - it should normally log > *something*. There is > nothing from postgresql in the eventlog? > > > > Based on what I've gleaned on reading some of the > past mail > > archives, I thought that this should > > work. Have I misunderstood something? > > It should,m as long as they are on the same version. > And it should > absolutely log an error. > > Can you try starting the database manually using > pg_ctl (not as a > service) and see if you get more errors then? > Remember to start it from > a runas:ed shell, so you run it as the same user. > > //Magnus > > ---------------------------(end of > broadcast)--------------------------- > TIP 5: don't forget to increase your free space map > settings > ____________________________________________________ Start your day with Yahoo! - make it your home page http://www.yahoo.com/r/hs
Did you check file ownership and permissions? Greetings. -----Mensaje original----- De: pgsql-admin-owner@postgresql.org [mailto:pgsql-admin-owner@postgresql.org]En nombre de Magnus Hagander Enviado el: lunes, 15 de agosto de 2005 16:07 Para: Dirk Kalp; pgsql-admin@postgresql.org Asunto: Re: [ADMIN] Server Won't Start On a Copy of Its Own Data Folder > Server Won't Start On a Copy of Its Own Data Folder > > 1. Installed the latest 8.0.3 binary on a Windows XP Pro > system as a service to be started on bootup. > 2. Used psql just to verify that I can connect. > 3. Used Windows Control Panel.Administrative Tools.Services > to stop the server cleanly. > 4. Made a copy of the PostgreSQL/8.0/data folder > 5. Renamed the installed folder as PostgreSQL/8.0/saved-data > 6. Renamed the copy of data as PostrgeSQL/8.0/data > 7. Used Windows Control Panel.Administrative Tools.Services > to start the server > 8. The server fails to come up > > Windows comes back and tells me that the service started and > stopped. The Windows > EventLog show that the postgres server was sent a start > control, entered the running state and > then entered the stopped state. There was nothing new in the > pg_log folder to provide > any clue. This is very weird - it should normally log *something*. There is nothing from postgresql in the eventlog? > Based on what I've gleaned on reading some of the past mail > archives, I thought that this should > work. Have I misunderstood something? It should,m as long as they are on the same version. And it should absolutely log an error. Can you try starting the database manually using pg_ctl (not as a service) and see if you get more errors then? Remember to start it from a runas:ed shell, so you run it as the same user. //Magnus ---------------------------(end of broadcast)--------------------------- TIP 5: don't forget to increase your free space map settings