Hitoshi Mitake

@imrook The original motivation of the journaling was improving performance. In the context of sheepdog, improve data integrity with journaling is impossible. For example, we cannot update multiple inode objects...

It is hard to implement dedup because it changes relation between inodes and objects. There are bunch of things if you want to implement it (I cannot list up soon)....

Yes, offline dedup (or its dry run mode) would be a good start point if we really work on it.

No, it isn't supported. You need to export the VDI and import it again.

Of course the compression will introduce tradeoff between cpu usage and disk usage. I think implementing a dry run command, which calculates a degree of disk resource reduction, will be...

@willbryant thanks for filing. The feature must be useful. Could you work on it, @yamada-h or @tmenjo ? If you don't have a time for it, I'll work on it...

@gangzai thanks for your report. However, the cluster snapshot feature is not maintained and probably we'll remove it in the feature. Because "storing every vdi in a single drive" is...

@gangzai yes, the snapshot is maintained. Could you provide entire command line you executed?

Hi @hmex , sorry for my late reply. Please don't use object cache and journaling. They are broken and don't improve performance. For consistency checking, `dog vdi check` can be...

@hmex sorry for inconvenience. I agree with deleting the features. Especially object cache is harmful because of not only its buggy implementation but also its broken design. The reason why...