CURP - fast replication nearly for free and without the current caveats about keydb

[JohnSully]

It looks very interesting. Yes i will be investigating this.

[dumblob]

Did you have some time to take a look at that paper?

[JohnSully]

@dumblob Yes I have looked, but its not quite time to implement this for KeyDB just yet.

Right now the focus for KeyDB is maximizing vertical scaling. In support of that we have MVCC coming out this week. Once we’ve maximized what a single node can do I will start to take a deeper look into CDRT/CURPS for active replication.

[dumblob]

@JohnSully understood, thanks for the update. We're considering switching to cloud based international provider of Redis API (does KeyDB.dev or EQ ALPHA provide such services?), but with way better scalability (including failover/HA), throughput and pushed limits (e.g. limitation by RAM) while guaranteeing full durability (we currently use appendonly yes and appendfsync always).

Yet another requirement on the chosen technology is the possibility to run everything as an on-premise DB (without any restrictions) for some special customers (in other words the chosen technology has to be open source without any limitations for commercial use).