Module11_Backup_PITR

Question 1

what are these PG_DUMP options
-a
-s
-n
-t
-f
-j njobs
-B

Answer 1

-a DATA ONLY (no schema deffintions)
-s NO DATA only schema
-n Specific schema only
-t Specific table only
-f FILE_NAME
-j njobs: parallel dumps
-B exclude blobs

Question 2

how to zip dump during creation

Answer 2

pg_dump DBNAME | gzip > FILENAME.gz

Question 3

how to take a logical backup of all cluster

Answer 3

pg_dumpall > filename.backup

Question 4

when to use pg_restore

Answer 4

when dump is not plain text or you want more features like import a single table only or only objects from 1 schema.

Question 5

how to configure continues archiving of wall files

Answer 5

WAL_LEVEL must be replica
Archive_mode must be ON
archive_command must be set
restart the server

Question 6

why would you use streaming wall

Answer 6

you don’t have to wait for the switch of the wall it automatically writes the change to file right away

Question 7

how to take offline backup from server

Answer 7

stop PG Services and copy data directory

Question 8

what is prefrerred methode for online backup

Answer 8

pg_basebackup -h localhost -D /DATA_DIR/

Question 9

how to check if pg_basebackup is correct

Answer 9

with pg_verifybackup tool.

Question 10

how to peform a PITR

Answer 10

• Stop the server and take a filesystem backup and clean the data directory
• copy files from backup to data directory with cp -rp
• configure recovery setting to postgresql.conf and create recovery.signal
• start server

Recovery settings that must be configured are:
Restore_command

maybe also:
Recovery_target_time