Is there a way to create a table in SS that uses both a normal table and also an Object Store? And set criteria based on which data is rolled over to OS
Hi Sandy. Please take a look at this for details about how unlimited storage databases work:
A database is either a regular database or an unlimited storage database. Only the latter are backed up by object storage.
There’s really no need to set criteria about what data gets saved to object storage. All data will go there within a couple of minutes of creation. But transaction commits are local to the cluster and updates are much quicker that way. We can commit a transaction in much less time than it takes to write to an external object store, like S3.
Recently-accessed data will be cached local on the cluster and we don’t have to go to the object store to get it.