most advanced settings are not in UI, and some UI areas require a lot of rework (exclusion patters, restore workflow) outstanding suppprt where the developer responds quickly and fixes issues users face even while using the products in most weird and unsupported scenarios (I’m one of those users) has stable version and exists for a long time. out of the box support for excluding known transient files (such as TimeMachine exclusion by xattr) - this drastically reduces the need to write custom exclusion patters. super fast exclusion patters, including regex rather simple to make it work with immutable storage support for asymmetric RSA encryption (different backup and restore keys) tons of cloud destinations supported, including google drive erasure coding making possible use of ureliable storage (hard drives) as backup destination a Generally, not very friendly support forum - this is subjective of course. At which point will he throw in the towel? The dude works at google and develops it on the weekends. unclear future and support model - most questions and requests are answered by “here is source, pull requests are welcome”. Can be workaround with rclone destination but this adds a layer of complexity). very limited support for cloud destinations (no *Drive type services. no stable version yet - so not suitable for production rather complex multilevel datastore structure that if corrupted not clear how to recover. My suggestion to implement that on the forums was met with complete silence. does not support extended attributes driven exclusions (for example to exclude all Files skipped by Time Machine on a mac).

support to run server side component to facilitate ACLs.It’s nice and frequently requested feature on duplicacy forums. ability to mount backup history as a virtual drive.backup configuration is stored on the destination.Generally, on paper, there are a lot of compelling features that build on what duplicacy does and move that forward. I admit, it could have been my fault, but that never happened with duplicacy, and it should not have been that easy to accomplish for a novice user.Īpproach to deduplication is indeed similar. Only compatibility (migration from) with previous version is promised, and as expected with software version 0.xxx any breaking changes can happen any time.ĭuring my time trying Kopia with local SFTP storage I managed to corrupt the datastore twice, both cases failed to recover. a set of snapshots), for example - maybe it's not even possible?) and for all its pretty graphs, Duplicacy doesn't really tell me anything useful either.Ĭontext: long time Duplicacy user here, and tried Kopia recently for a month.įor production use Kopia is not even a contender - it is still alpha. Unfortunately both products have so little status information about what they're up to that it's fairly hard to quickly compare this stuff (I'm yet to figure out how to even get the entire size of a backup set in Kopia (i.e. Duplicacy makes a big fuss about being the first and only backup program to de-duplicate all backup sets sent to the same destination (lock free deduplication they call it) - but as far as I can tell, Kopia will also do this? Is this correct? Anyone know how they compare in this regard?

Specifically, I'm trying to figure out how their de-duplication and compression performance compares (when using multiple backup sets to the same destination). The UI's and the way they're set up to work from a workflow perspective is quite different, but I feel like behind the scenes the backup mechanism is probably similar? Many people seem to use Duplicacy, but it's quite hard to find info about Kopia.

Duplicacy vs duplicity