Replication between units

I checked the docs and did not see and answer to this one.

Is replication block level or file level?

Im setting up my second rockstor NAS (SSD’s) and wondering if I can replicate my VM images. If it is block level it should be fast, but if it is file level then it will take a few min per image to replicate the large file.


@clink Hello again.

Roughly speaking it’s more block level.

Hope that helps.

Please explain in more detail?

Im interested in how that works

@clink I’m afraid an explanation from me would be redundant as we essentially ‘wrap’ btrfs send/receive. At least in the context of your question re level of replication function.

So do a search on btrfs send/receive and you should have your answer.

Our wrapper does the nice-to-have snapshot management around this send/receive mechanism but does nothing more than makes it easier to use. Your question relates primarily to the lower level function of btrfs which is not Rockstor specific.

Hope that helps.

1 Like

Thank you that helps. I did not know what mechanism was being used. With this I can dig in and find the information Im looking for.

1 Like

@clink No worries.
This would be a good time to create another rockstor-doc GitHub issue if you fancy:

I believe we have one or two already re improving the replication docs which are definitely pretty weak at the moment. If you find a relevant one then please add that we should stress it’s send/receive based. That way we can hopefully be proactive in removing the questions that are easily missed once one is more familiar with the system. Which the main developers are more likely to be.

Just a thought.