Thread: Error while upgrading from 8.4 to 9.3
Hi, we are trying to upgrade from 8.4 to 9.3, but at the end of the process we've got the error: command: "/usr/lib/postgresql/9.3/bin/pg_resetxlog" -o 2429057 "/mnt/datos/datos_9.3/main" >> "pg_upgrade_utility.log" 2>&1 pg_resetxlog: lock file «postmaster.pid» exists Is a server running? If not, delete the lock file and try again. Our setup is this: Server: Ubuntu 12.04 – 32 bits Clusters: Postgresql 8.4 y Postgresql 9.3 The steps we followed are: 1) pg_upgrade using –-check result Ok. 2) pg_upgrade without --check Everything went ok, data directory was copied without problems. At the end, we got the error I mentioned at the beggining of this email. P.S.: pg_upgrade_utility.log attached. Regards, -- Leonardo M. Ramé Medical IT - Griensu S.A. Av. Colón 636 - Piso 8 Of. A X5000EPT -- Córdoba Tel.: +54(351)4246924 +54(351)4247788 +54(351)4247979 int. 19 Cel.: +54 9 (011) 40871877
Attachment
On 05/28/2014 06:19 AM, Leonardo M. Ramé wrote: > Hi, we are trying to upgrade from 8.4 to 9.3, but at the end of the > process we've got the error: > > command: "/usr/lib/postgresql/9.3/bin/pg_resetxlog" -o 2429057 "/mnt/datos/datos_9.3/main" >> "pg_upgrade_utility.log"2>&1 > pg_resetxlog: lock file «postmaster.pid» exists > Is a server running? If not, delete the lock file and try again. > > Our setup is this: > Server: Ubuntu 12.04 – 32 bits > Clusters: Postgresql 8.4 y Postgresql 9.3 > > The steps we followed are: > > 1) pg_upgrade using –-check result Ok. > 2) pg_upgrade without --check Can you show us the complete command with arguments for both 1 & 2 above? > > Everything went ok, data directory was copied without problems. > > At the end, we got the error I mentioned at the beggining of this email. > > P.S.: pg_upgrade_utility.log attached. > > Regards, > > > > -- Adrian Klaver adrian.klaver@aklaver.com
On Wed, May 28, 2014 at 10:19:24AM -0300, Leonardo M. Ramé wrote: > Hi, we are trying to upgrade from 8.4 to 9.3, but at the end of the > process we've got the error: > > command: "/usr/lib/postgresql/9.3/bin/pg_resetxlog" -o 2429057 "/mnt/datos/datos_9.3/main" >> "pg_upgrade_utility.log"2>&1 > pg_resetxlog: lock file «postmaster.pid» exists > Is a server running? If not, delete the lock file and try again. > > Our setup is this: > Server: Ubuntu 12.04 – 32 bits > Clusters: Postgresql 8.4 y Postgresql 9.3 > > The steps we followed are: > > 1) pg_upgrade using –-check result Ok. > 2) pg_upgrade without --check > > Everything went ok, data directory was copied without problems. > > At the end, we got the error I mentioned at the beggining of this email. > > P.S.: pg_upgrade_utility.log attached. I am not sure what would cause this. At the point where you get this error, pg_upgrade has already started and stopped the old and new servers, so I am not sure why the pid file would exist. Is it possible the file copy was so fast that the new cluster had not fully shut down? Seems odd because 9.3 has code to detect that the server is fully down. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + Everyone has their own god. +
On 05/28/2014 07:24 AM, Claudio Biasatti wrote: > with check > > /usr/lib/postgresql/9.3/bin/pg_upgrade --check -b > /usr/lib/postgresql/8.4/bin/ -B /usr/lib/postgresql/9.3/bin/ -d > /mnt/datos/ -D /mnt/datos/datos_9.3/main/ -o ' -c > config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > config_file=/etc/postgresql/9.3/main/postgresql.conf' > > > without check > > /usr/lib/postgresql/9.3/bin/pg_upgrade -b /usr/lib/postgresql/8.4/bin/ > -B /usr/lib/postgresql/9.3/bin/ -d /mnt/datos/ -D > /mnt/datos/datos_9.3/main/ -o ' -c > config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > config_file=/etc/postgresql/9.3/main/postgresql.conf' > > So what are the port numbers in the conf files. I have not used that option, but is there a chance you are double starting the servers? Starting an instance for each using the port numbers from the conf files and then pg_upgrade is starting them on 50432. -- Adrian Klaver adrian.klaver@aklaver.com
On 2014-05-28 07:39:33 -0700, Adrian Klaver wrote: > On 05/28/2014 07:24 AM, Claudio Biasatti wrote: > >with check > > > >/usr/lib/postgresql/9.3/bin/pg_upgrade --check -b > >/usr/lib/postgresql/8.4/bin/ -B /usr/lib/postgresql/9.3/bin/ -d > >/mnt/datos/ -D /mnt/datos/datos_9.3/main/ -o ' -c > >config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > >config_file=/etc/postgresql/9.3/main/postgresql.conf' > > > > > >without check > > > >/usr/lib/postgresql/9.3/bin/pg_upgrade -b /usr/lib/postgresql/8.4/bin/ > >-B /usr/lib/postgresql/9.3/bin/ -d /mnt/datos/ -D > >/mnt/datos/datos_9.3/main/ -o ' -c > >config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > >config_file=/etc/postgresql/9.3/main/postgresql.conf' > > > > > > So what are the port numbers in the conf files. > > I have not used that option, but is there a chance you are double > starting the servers? > > Starting an instance for each using the port numbers from the conf > files and then pg_upgrade is starting them on 50432. > > Sorry guys, one of the in-site tech support guys started the postgresql service while we were trying to do the upgrade. Looks like this was the cause of the failure. Regards, -- Leonardo M. Ramé Medical IT - Griensu S.A. Av. Colón 636 - Piso 8 Of. A X5000EPT -- Córdoba Tel.: +54(351)4246924 +54(351)4247788 +54(351)4247979 int. 19 Cel.: +54 9 (011) 40871877
On Wed, May 28, 2014 at 12:05:54PM -0300, Leonardo M. Ramé wrote: > On 2014-05-28 07:39:33 -0700, Adrian Klaver wrote: > > On 05/28/2014 07:24 AM, Claudio Biasatti wrote: > > >with check > > > > > >/usr/lib/postgresql/9.3/bin/pg_upgrade --check -b > > >/usr/lib/postgresql/8.4/bin/ -B /usr/lib/postgresql/9.3/bin/ -d > > >/mnt/datos/ -D /mnt/datos/datos_9.3/main/ -o ' -c > > >config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > > >config_file=/etc/postgresql/9.3/main/postgresql.conf' > > > > > > > > >without check > > > > > >/usr/lib/postgresql/9.3/bin/pg_upgrade -b /usr/lib/postgresql/8.4/bin/ > > >-B /usr/lib/postgresql/9.3/bin/ -d /mnt/datos/ -D > > >/mnt/datos/datos_9.3/main/ -o ' -c > > >config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c > > >config_file=/etc/postgresql/9.3/main/postgresql.conf' > > > > > > > > > > So what are the port numbers in the conf files. > > > > I have not used that option, but is there a chance you are double > > starting the servers? > > > > Starting an instance for each using the port numbers from the conf > > files and then pg_upgrade is starting them on 50432. > > > > > > Sorry guys, one of the in-site tech support guys started the postgresql > service while we were trying to do the upgrade. Looks like this was the > cause of the failure. OK. We block users from connecting to our own postmaster by placing the socket files in the current directory, but that doesn't prevent someone from starting the server using the default configuration. Not sure how we could block that. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + Everyone has their own god. +
with check /usr/lib/postgresql/9.3/bin/pg_upgrade --check -b /usr/lib/postgresql/8.4/bin/ -B /usr/lib/postgresql/9.3/bin/ -d /mnt/datos/ -D /mnt/datos/datos_9.3/main/ -o ' -c config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c config_file=/etc/postgresql/9.3/main/postgresql.conf' without check /usr/lib/postgresql/9.3/bin/pg_upgrade -b /usr/lib/postgresql/8.4/bin/ -B /usr/lib/postgresql/9.3/bin/ -d /mnt/datos/ -D /mnt/datos/datos_9.3/main/ -o ' -c config_file=/etc/postgresql/8.4/main/postgresql.conf' -O ' -c config_file=/etc/postgresql/9.3/main/postgresql.conf' ----Mensaje original----- From: Adrian Klaver Sent: Wednesday, May 28, 2014 11:15 AM To: "Leonardo M. Ramé" ; PostgreSql-general Cc: c.biasatti@griensu.com Subject: Re: [GENERAL] Error while upgrading from 8.4 to 9.3 On 05/28/2014 06:19 AM, Leonardo M. Ramé wrote: > Hi, we are trying to upgrade from 8.4 to 9.3, but at the end of the > process we've got the error: > > command: "/usr/lib/postgresql/9.3/bin/pg_resetxlog" -o 2429057 > "/mnt/datos/datos_9.3/main" >> "pg_upgrade_utility.log" 2>&1 > pg_resetxlog: lock file «postmaster.pid» exists > Is a server running? If not, delete the lock file and try again. > > Our setup is this: > Server: Ubuntu 12.04 – 32 bits > Clusters: Postgresql 8.4 y Postgresql 9.3 > > The steps we followed are: > > 1) pg_upgrade using –-check result Ok. > 2) pg_upgrade without --check Can you show us the complete command with arguments for both 1 & 2 above? > > Everything went ok, data directory was copied without problems. > > At the end, we got the error I mentioned at the beggining of this email. > > P.S.: pg_upgrade_utility.log attached. > > Regards, > > > > -- Adrian Klaver adrian.klaver@aklaver.com
On 05/28/2014 08:15 AM, Bruce Momjian wrote: > On Wed, May 28, 2014 at 12:05:54PM -0300, Leonardo M. Ramé wrote: >> >> Sorry guys, one of the in-site tech support guys started the postgresql >> service while we were trying to do the upgrade. Looks like this was the >> cause of the failure. > > OK. We block users from connecting to our own postmaster by placing the > socket files in the current directory, but that doesn't prevent someone > from starting the server using the default configuration. Not sure how > we could block that. > So for my edification, does that mean -O ' -c config_file=/etc/postgresql/9.3/main/postgresql.conf' could cause a problem or that starting Postgres outside pg_upgrade could cause a problem, or both. -- Adrian Klaver adrian.klaver@aklaver.com
On Wed, May 28, 2014 at 04:56:53PM -0700, Adrian Klaver wrote: > On 05/28/2014 08:15 AM, Bruce Momjian wrote: > >On Wed, May 28, 2014 at 12:05:54PM -0300, Leonardo M. Ramé wrote: > > >> > >>Sorry guys, one of the in-site tech support guys started the postgresql > >>service while we were trying to do the upgrade. Looks like this was the > >>cause of the failure. > > > >OK. We block users from connecting to our own postmaster by placing the > >socket files in the current directory, but that doesn't prevent someone > >from starting the server using the default configuration. Not sure how > >we could block that. > > > > So for my edification, does that mean -O ' -c > config_file=/etc/postgresql/9.3/main/postgresql.conf' could cause a > problem or that starting Postgres outside pg_upgrade could cause a > problem, or both. Starting Postgres outside pg_upgrade would _certainly_ cause a problem. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + Everyone has their own god. +
Bruce Momjian <bruce@momjian.us> writes:
> On Wed, May 28, 2014 at 04:56:53PM -0700, Adrian Klaver wrote:
>> So for my edification, does that mean -O ' -c
>> config_file=/etc/postgresql/9.3/main/postgresql.conf' could cause a
>> problem or that starting Postgres outside pg_upgrade could cause a
>> problem, or both.
> Starting Postgres outside pg_upgrade would _certainly_ cause a problem.
It appeared to me that it was failing safely, ie, once somebody had
started a conflicting postmaster, pg_upgrade's attempt to start one
failed without damage.
regards, tom lane
On Wed, May 28, 2014 at 11:28:47PM -0400, Tom Lane wrote: > Bruce Momjian <bruce@momjian.us> writes: > > On Wed, May 28, 2014 at 04:56:53PM -0700, Adrian Klaver wrote: > >> So for my edification, does that mean -O ' -c > >> config_file=/etc/postgresql/9.3/main/postgresql.conf' could cause a > >> problem or that starting Postgres outside pg_upgrade could cause a > >> problem, or both. > > > Starting Postgres outside pg_upgrade would _certainly_ cause a problem. > > It appeared to me that it was failing safely, ie, once somebody had > started a conflicting postmaster, pg_upgrade's attempt to start one > failed without damage. Yes, it was. The thing we can't block is an administrator starting the old or new postmaster while pg_upgrade isn't using it, then shutting it down before pg_upgrade needs it --- seems unlikely though. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + Everyone has their own god. +