Обсуждение: On Replica - History issue
Hello I need some advice about the issue in the bellow. Because of datacenter I has to create my replica's from zero. On a cluster I get a error, it's look like about pgbackrest issue, but also this one has no backup for now. Before it has. Is there anything I should do Manuel?
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.810 +03 [316956] LOG: startup process (PID 316957) exited with exit code 1
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.810 +03 [316956] LOG: aborting startup due to startup process failure
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.811 +03 [316956] LOG: database system is shut down
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: pg_ctl: could not start server
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: Examine the log output.
Oct 27 13:33:26 se-prd-db02 systemd[1]: postgresql@13-clustername.service: Can't open PID file /run/postgresql/13-clustername.pid (yet?) after start: Operation not permitted
Oct 27 13:33:26 se-prd-db02 systemd[1]: postgresql@13-clustername.service: Failed with result 'protocol'.
Oct 27 13:33:26 se-prd-db02 systemd[1]: Failed to start PostgreSQL Cluster 13-clustername.
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.810 +03 [316956] LOG: startup process (PID 316957) exited with exit code 1
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.810 +03 [316956] LOG: aborting startup due to startup process failure
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.811 +03 [316956] LOG: database system is shut down
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: pg_ctl: could not start server
Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: Examine the log output.
Oct 27 13:33:26 se-prd-db02 systemd[1]: postgresql@13-clustername.service: Can't open PID file /run/postgresql/13-clustername.pid (yet?) after start: Operation not permitted
Oct 27 13:33:26 se-prd-db02 systemd[1]: postgresql@13-clustername.service: Failed with result 'protocol'.
Oct 27 13:33:26 se-prd-db02 systemd[1]: Failed to start PostgreSQL Cluster 13-clustername.
On Fri, 2023-10-27 at 13:36 +0300, SOzcn wrote: > Hello I need some advice about the issue in the bellow. Because of datacenter I has > to create my replica's from zero. On a cluster I get a error, it's look like about > pgbackrest issue, but also this one has no backup for now. Before it has. Is there > anything I should do Manuel? > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restorefile "00000003.history" from archive: command not found I don't think Manuel is at fault. You should use the absolute path to "pgbackrest" in the "restore_command". Yours, Laurenz Albe
Hello,
"restore_command" is default .
my archive command is ;
archive_command = 'pgbackrest --stanza=cluster_stanza archive-push %p'
with this way I setup my backup procedure without problem, is there might be another problem I cannot see?
Laurenz Albe <laurenz.albe@cybertec.at>, 27 Eki 2023 Cum, 15:30 tarihinde şunu yazdı:
On Fri, 2023-10-27 at 13:36 +0300, SOzcn wrote:
> Hello I need some advice about the issue in the bellow. Because of datacenter I has
> to create my replica's from zero. On a cluster I get a error, it's look like about
> pgbackrest issue, but also this one has no backup for now. Before it has. Is there
> anything I should do Manuel?
>
> Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found
> Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
I don't think Manuel is at fault.
You should use the absolute path to "pgbackrest" in the "restore_command".
Yours,
Laurenz Albe
I also checked my prod cluster wal file, there has no that file either. I missing something I don't see.
SOzcn <selahattinozcnma@gmail.com>, 30 Eki 2023 Pzt, 13:48 tarihinde şunu yazdı:
Hello,"restore_command" is default .my archive command is ;archive_command = 'pgbackrest --stanza=cluster_stanza archive-push %p'with this way I setup my backup procedure without problem, is there might be another problem I cannot see?Laurenz Albe <laurenz.albe@cybertec.at>, 27 Eki 2023 Cum, 15:30 tarihinde şunu yazdı:On Fri, 2023-10-27 at 13:36 +0300, SOzcn wrote:
> Hello I need some advice about the issue in the bellow. Because of datacenter I has
> to create my replica's from zero. On a cluster I get a error, it's look like about
> pgbackrest issue, but also this one has no backup for now. Before it has. Is there
> anything I should do Manuel?
>
> Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found
> Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
I don't think Manuel is at fault.
You should use the absolute path to "pgbackrest" in the "restore_command".
Yours,
Laurenz Albe
On Mon, 2023-10-30 at 13:48 +0300, SOzcn wrote: > > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could notrestore file "00000003.history" from archive: command not found > > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found > > > > You should use the absolute path to "pgbackrest" in the "restore_command". > > "restore_command" is default . > my archive command is ; > archive_command = 'pgbackrest --stanza=cluster_stanza archive-push %p' > > with this way I setup my backup procedure without problem, is there might be another problem I cannot see? You get a fatal error restoring a file from the archive, because the restore command is not found. What is unclear about the error message? Why don't you try the absolute path? Yours, Laurenz Albe
I do with absolute path already, that's the problem.
Laurenz Albe <laurenz.albe@cybertec.at>, 30 Eki 2023 Pzt, 15:29 tarihinde şunu yazdı:
On Mon, 2023-10-30 at 13:48 +0300, SOzcn wrote:
> > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
> > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found
> >
> > You should use the absolute path to "pgbackrest" in the "restore_command".
>
> "restore_command" is default .
> my archive command is ;
> archive_command = 'pgbackrest --stanza=cluster_stanza archive-push %p'
>
> with this way I setup my backup procedure without problem, is there might be another problem I cannot see?
You get a fatal error restoring a file from the archive, because the restore command is not found.
What is unclear about the error message? Why don't you try the absolute path?
Yours,
Laurenz Albe
Can you give me more details please? I've been created more than 20 cluster replica without problem, on a cluster I get this and unfortunately I can't figure it out like as I say on top.
SOzcn <selahattinozcnma@gmail.com>, 30 Eki 2023 Pzt, 16:27 tarihinde şunu yazdı:
I do with absolute path already, that's the problem.Laurenz Albe <laurenz.albe@cybertec.at>, 30 Eki 2023 Pzt, 15:29 tarihinde şunu yazdı:On Mon, 2023-10-30 at 13:48 +0300, SOzcn wrote:
> > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: 2023-10-27 13:33:26.808 +03 [316957] FATAL: could not restore file "00000003.history" from archive: command not found
> > > Oct 27 13:33:26 se-prd-db02 postgresql@13-clustername[316944]: sh: 1: pgbackrest: not found
> >
> > You should use the absolute path to "pgbackrest" in the "restore_command".
>
> "restore_command" is default .
> my archive command is ;
> archive_command = 'pgbackrest --stanza=cluster_stanza archive-push %p'
>
> with this way I setup my backup procedure without problem, is there might be another problem I cannot see?
You get a fatal error restoring a file from the archive, because the restore command is not found.
What is unclear about the error message? Why don't you try the absolute path?
Yours,
Laurenz Albe