moving backups to s3

[liorme]

hi. in a scaleout repo with s3, why is it only after i do a second active full (meaning : 1 full + let's say 3 increments + active full) does the first 1+3 increments go to the cloud? can't veeam upload everything to the cloud from the start? and i mean move, not copy.

thank you

[sfirmes]

The “move mode” only moves inactive backup chains to the S3 target. This is the behavior you are seeing. The “copy mode” will copy the backup data to the S3 target once the backup file in the performance tier is created. You can use both modes together as well.

[liorme]

thank you. if i use both modes together, it would take twice the space, no?

[Mildur]

No, space is needed only once for the same restore point/blocks.

[liorme]

cool. so actually, let's say i have a first full backup, and then some restore points after it - and i have a copy and move enabled.
so the first full and it's points, will be copied to the cloud.
now, after some time, the next active full is created, and now, the veeam will move the first full and it's point to the cloud and only leaves a small metadata behind. will this move action not take more space? the block storage will "trace" the identical blocks and will not "eat" the storage twice?

[oleg.feoktistov]

Correct, if your inactive chain was moved with copy mode already, when offload policy hits, blocks from inactive chain are removed from performance tier and just metadata remains. Only unique blocks are moved to capacity tier, if any. Thanks!

[liorme]

cool, thank you