chunk commit process during backup job using ddb

  • 22 March 2021
  • 1 reply

Badge +2

Hi all,

Is it possible if some explain the process of committing chunks to lib and signature in ddb. I read somewhere that chunk will keep on noting all data blocks even it is repeated. Actually m confused with that statement . Might b my understanding is wrong but need help to understand exactly when ddb signatures will be committed to ddb and when will chunk will get committed ?


Best answer by Damian Andre 22 March 2021, 21:53

View original

1 reply

Userlevel 7
Badge +21

Hey @Karamveer Singh 


I think this video may answer a lot of your questions:


But the simple version is, if its a new block we’ve never seen before, we store the data in the chunk and comit the signature in the DDB. If its a block we have encountered before, we don’t store the data, but we put a link to where the data is in the chunk exists instead. Think of it like a shortcut or symbolic link. This is why you don’t need the deduplication database to exist to perform restores - the chunks link to the data rather than having to query the database to find out where the data is.