Hi @james4,
When a new set of data is added to odrive’s scope, it will scan through the entire set and try to determine the most efficient way to upload. If the number of objects is very large this can take a very long time, and it can increase processing overhead to the point that you start getting some resource exhaustion.
To avoid this, I always recommend importing data in segments, instead of all in one go. This reduces the overhead and allows odrive to process the subset of data faster. If you can break the set up into smaller chunk and wait until a chunk has finished before adding another you should see better results.
Additionally, if there is already a lot of data that odrive is tracking, the new data is additive. odrive still tracks all of the existing data, on both side, to determine if any changes have been made. This will also increase overhead as the dataset grows and grows. If you haven’t already done so, you can unsync portions of the data that have already synced up and that you do not need immediate access to to reduce the amount of work odrive has to perform continually. This will also help move things along for new imports.