Have you also considered running a read replica off your node that runs api.primersion.com and then just using that to take the snapshots so you don’t need an as powerful server?
But I would still have to shut down the node for taking the snapshot, so the data does not change in the meantime, right? Or is it possible to pause replication?
Is that the snapshot or the actual node itself?
Haven't really tracked the node size so far, but I'm guessing it's 1.5 to 2 times as much. Probably depends on compression level and stuff.
Would be interesting data for sure.
Have you also considered running a read replica off your node that runs api.primersion.com and then just using that to take the snapshots so you don’t need an as powerful server?
But I would still have to shut down the node for taking the snapshot, so the data does not change in the meantime, right? Or is it possible to pause replication?
Ah yeah I forgot that's needed. Might be a way to pause replication, I don't work with mongo enough to know the answer of the top of my head.