Tomorrow will be the second anniversary of the only info about Rockstor replication (http://goo.gl/rC0wCL). This pre-dates v3.8.10 which arrived two months ago and was a ground-up rewrite. Some questions:
- What is the rationale behind many replications being done (and snapshots created on the receiver) before a share is visible?
- Is data within snapshots on receiver accessible before this? If so, how?
- When share appears after the 6th(?) replication, do subsequent replications merge with it, so if you wanted to restore, you would for example, create a samba export of the share and the new data would be there? IE, you don’t have to convert the latest snapshot into a share and export that.
- If the initial rep is a bulk mirror copy, and all subsequent reps are incrementals ad infinitum, eventually the receiver will run out of space. Can you delete older snaps, or will this confuse\break replication?
- Is there any alternative to doing a bulk initial replication by WAN? After a local rep, shipping the disk\s to destination, and importing the pool, how can the receiver NAS be manually coerced to resume replication?
I think a blog would be very helpful. Many thanks.