Hi everybody. I’ve tested many Open Source products for creating backup in scenario described in Subject.
I found two: Duplicati and Ferrobackup do that well in file mode(only have this mode). It means that in incemental backup there are only changed fragments of files (opened too, because of vss snapshots), so finally dayli backup files they are small.
My scenario was launching incemental backup two times, when the source files are not changed (I’ve shat down SQL SERVER, and I did not launch MS OUTLOOK)
When I tried to do that scenario in urbackup in file level each copies has 2.5 GB like whole .pst files, so this variants is disappointing. It’s suprising, because in docs is written, that urbackup archives only changed fragment of source files, comapring hashes of files.
In block level scenario situations is better, but not exellent - every incremental copies of volume C: have 500 - 600 MB (.pst file has 2.5 GB, and databases has 16 GB ), so conclusion is, that backups takes some fragments of files, but why it so big?. Maybe some Temp files changes fragment of volume C? In I’ve compared Duplicati and Ferrobackup incemental backups -backups have 5 or 10 MB of real changed fragment of files.
Any ideas, where is the reason ?