Differences

This shows you the differences between two versions of the page.

--- postgresql_pgbackrest_configuration [2025/05/17 15:24] – andonovj
+++ postgresql_pgbackrest_configuration [2025/05/19 06:22] (current) – [Restore] andonovj
@@ Line 15: / Line 15: @@
 </Code>
-So in nutshell, my patroni.yml looks like this:
-<Code:bash|Patroni Yaml>
+After that, you can configure the /etc/pgbackrest.conf file on all servers: db nodes + backup server
-scope: stampede
-name: ${host}
-restapi:
-  listen: ${host}:8008
-  connect_address: ${host}:8008
-etcd:
-  hosts: etcd00:2379, etcd01:2379, etcd02:2379
-bootstrap:
-  dcs:
-    ttl: 30
-    loop_wait: 10
-    retry_timeout: 10
-    maximum_lag_on_failover: 1048576
-    maximum_lag_on_syncnode: 15000000
-    synchronous_mode: true
-    postgresql:
-      use_pg_rewind: true
-      use_slots: true
-  initdb:
+<Code:bash|Pgbackrest Configuration>
-  - encoding: UTF8
+[global]
-  - data-checksums
+repo1-path=/backups
+repo1-retention-full=14
+repo1-retention-full-type=time
+repo1-host-user=pgbackrest
+archive-check=n
+process-max=2
+log-level-console=info
+log-path=/backup/pgbackrest/log
+log-level-file=debug
+start-fast=y (performs a checkpoint)
+delta=y
-  pg_hba:
+[cluster_backup]
-  - host replication rep_user ${subnet}.0/24 md5
+pg1-path=/db/pgdata
-  - host all all ${subnet}.0/24 md5
-postgresql:
-  listen: ${host}:5432
-  connect_address: ${host}:5432
-  data_dir: /db/pgdata
-  bin_dir: ${bindir}
-  pgpass: /tmp/pgpass0
-  authentication:
-    replication:
-      username: rep_user
-      password: newpass
-    superuser:
-      username: postgres
-      password: newpass
-  parameters:
-    unix_socket_directories: '/var/run/postgresql'
-    external_pid_file: '/var/run/postgresql/17-main.pid'
-    logging_collector: "on"
-    log_directory: "/var/log/postgresql"
-    log_filename: "postgresql-17-main.log"
-    shared_buffers: 100MB
-    work_mem: 16MB
-    maintenance_work_mem: 10MB
-    max_worker_processes: 16
-    wal_buffers: 16MB
-    max_wal_size: 200MB
-    min_wal_size: 100MB
-    effective_cache_size: 50MB
-    fsync: on
-    checkpoint_completion_target: 0.9
-    log_rotation_size: 100MB
-    listen_addresses: '*'
-    max_connections: 2000
-    temp_buffers: 4MB
-    archive_mode: "on"
-    wal_level: "replica"
-    archive_command: "pgbackrest --stanza=cluster_backup archive-push %p"
-    restore_command: "pgbackrest --stanza=cluster_backup archive-get %f %p"
 </Code>
@@ Line 173: / Line 122: @@
 Check if that is the correct state and if you are happy. Then promote it and THEN stop it:
+<Code:bash | Promote the PostgreSQL>
+/usr/pgsql-17/bin/pg_ctl promote -D /db/pgdata -w
+</Code>
 <Code:bash | stopping outside of Patroni>
@@ Line 188: / Line 141: @@
 That is how I was able to make it work. I am 99.(9)8% (Math nerds here), sure I do something wrong, but still. If you cannot get it work with patroni initially, that way works.
-=====Appendix=====
-Configuration for pgbackrest on ALL servers:
-<Code:bash|Configuration>
-[global]
-repo1-path=/backups
-repo1-retention-full=14
-repo1-retention-full-type=time
-repo1-host-user=pgbackrest
-archive-check=n
-process-max=2
-log-level-console=info
-log-path=/backup/pgbackrest/log
-log-level-file=debug
-start-fast=y (performs a checkpoint)
-delta=y
-[cluster_backup]
-pg1-path=/db/pgdata
-</Code>
-FYI: https://pgbackrest.org/configuration.html#section-general/