Database recovery from Text Backup in combination with Journal files (--new-autoupdate-with-journal)

Subject

When a valid database copy is not available for applying journal to, it is possible to recover data using a text backup file in combination with journal files, as long as the journal files period covers the date when the text backup was made.

Description of solution / instructions

1. Place the text backup as the only file in the Backup folder in format TBYYMMDD.TXT - date of the backup.
   1. the exactly correct file name is supposedly important, this is not known for sure
   2. if the backup starts before midnight, but finishes after midnight, the proper file name is the date before midnight.
2. Place the journal folder in its' place (make copies, this will get deleted after the process automatically)
   1. a known bug at the moment of writing - this process only supports a single journal file (use https://services.burti.lv/confluence/x/5Zq-BQ if necessary to merge multiple files into one)
3. Remove any pre-existing HANSA.HDB
4. Use the command line parameter --new-autoupdate-with-journal  - this will first perform a backup import (add db cache as necessary), then apply the journal immediately
5. The database will end up being corrupt - need to perform a raw data export (--rawdata)
6. Import the raw data backup

References

Additional information

Additional keywords

crash recovery journaling datubāze bojāta žurnāls žurnalēšana

Attachments