Question

我使用命令

每天备份postgresql数据库

/usr/bin/pg_basebackup -D $outdir -Ft -x -z -w -R -v

现在我想在另一台服务器上恢复此数据库。我使用了https://www.postgresql.org/docs/9.5/static/continuous-archiving.html#BACKUP-PITR-RECOVERY上的描述。

备份中包含的recovery.conf文件包含以下内容：

standby_mode = 'on'
primary_conninfo = 'user=postgres port=5432 sslmode=prefer sslcompression=1 krbsrvname=postgres'

文档中的下一步（8.）说启动postgresql。这会导致因超时而导致失败：

3783 postgres: startup process   waiting for 0000000100000024000000B

在原始服务器上，我没有此文件。是否可以在不使用任何WAL文件的情况下仅恢复pg_basebackup的状态？那么recovery.conf文件应该是什么？

根据@JosMac的建议，我使用以下结果移动了recovery.conf：

shaun2:/var/lib/pgsql/data # service postgresql start
Job for postgresql.service failed because the control process exited with error code. See "systemctl status postgresql.service" and "journalctl -xe" for details.
shaun2:/var/lib/pgsql/data # service postgresql status
â postgresql.service - PostgreSQL database server
   Loaded: loaded (/usr/lib/systemd/system/postgresql.service; enabled; vendor preset: disabled)
   Active: failed (Result: exit-code) since Mon 2018-06-18 12:02:53 CEST; 12s ago
  Process: 1340 ExecStop=/usr/lib/postgresql-init stop (code=exited, status=0/SUCCESS)
  Process: 9355 ExecStart=/usr/lib/postgresql-init start (code=exited, status=1/FAILURE)
 Main PID: 1060 (code=exited, status=0/SUCCESS)

Jun 18 12:02:52 shaun2 postgres[9369]: [3-1] 2018-06-18 12:02:52 CEST   LOG:  invalid checkpoint record
Jun 18 12:02:52 shaun2 postgres[9369]: [4-1] 2018-06-18 12:02:52 CEST   FATAL:  could not locate required checkpoint record
Jun 18 12:02:52 shaun2 postgres[9369]: [4-2] 2018-06-18 12:02:52 CEST   HINT:  If you are not restoring from a backup, try removing the file "/var/lib/pgsql/data/backup_label".
Jun 18 12:02:52 shaun2 postgres[9367]: [2-1] 2018-06-18 12:02:52 CEST   LOG:  startup process (PID 9369) exited with exit code 1
Jun 18 12:02:52 shaun2 postgres[9367]: [3-1] 2018-06-18 12:02:52 CEST   LOG:  aborting startup due to startup process failure
Jun 18 12:02:53 shaun2 postgresql-init[9355]: pg_ctl: could not start server
Jun 18 12:02:53 shaun2 systemd[1]: postgresql.service: Control process exited, code=exited status=1
Jun 18 12:02:53 shaun2 systemd[1]: Failed to start PostgreSQL database server.
Jun 18 12:02:53 shaun2 systemd[1]: postgresql.service: Unit entered failed state.
Jun 18 12:02:53 shaun2 systemd[1]: postgresql.service: Failed with result 'exit-code'.

我认为由于backup_label的内容，PostgreSQL仍然在寻找丢失的WAL文件：

shaun2:/var/lib/pgsql/data # cat backup_label
START WAL LOCATION: 24/B0000028 (file 0000000100000024000000B0)
CHECKPOINT LOCATION: 24/B0000028
BACKUP METHOD: streamed
BACKUP FROM: master
START TIME: 2018-06-14 02:55:08 CEST
LABEL: pg_basebackup base backup

将backup_label移走后的结果：

shaun2:/var/lib/pgsql/data # service postgresql status
â postgresql.service - PostgreSQL database server
   Loaded: loaded (/usr/lib/systemd/system/postgresql.service; enabled; vendor preset: disabled)
   Active: failed (Result: exit-code) since Mon 2018-06-18 12:17:54 CEST; 4s ago
  Process: 1340 ExecStop=/usr/lib/postgresql-init stop (code=exited, status=0/SUCCESS)
  Process: 10401 ExecStart=/usr/lib/postgresql-init start (code=exited, status=1/FAILURE)
 Main PID: 1060 (code=exited, status=0/SUCCESS)

Jun 18 12:17:53 shaun2 postgres[10414]: [4-1] 2018-06-18 12:17:53 CEST   LOG:  invalid secondary checkpoint record
Jun 18 12:17:53 shaun2 postgres[10414]: [5-1] 2018-06-18 12:17:53 CEST   PANIC:  could not locate a valid checkpoint record
Jun 18 12:17:54 shaun2 postgres[10412]: [2-1] 2018-06-18 12:17:54 CEST   LOG:  startup process (PID 10414) was terminated by signal 6: Aborted

Answer 1

我们使用pg_basebackup进行备份，并进行了多次修复，因此通常它可以正常运行而不会出现问题。

但我建议你使用参数-X stream而不是-x（意思是“-X fetch”）。使用此参数，pg_basebackup将捕获并存储在备份期间创建的WAL日志段以及数据文件。这些WAL日志将存储在单独的pg_xlog.tar或pg_wal.tar文件中（取决于PG版本）。

可以在此处找到完整的恢复说明 - pg_basebackup / pg-barman – restore tar backup

从pg_basebackup恢复

1 个答案: