Hi Team,
We have a setup in which we have an ODA as a target host to create the vDBs. The sources are diverse in terms of ODAs and Exadatas.
One of our Source DBs is huge, in the sense that the table not only has rows in the magnitude of billions but also columns in the order of hundreds. While normal DML operations on the vDB created from this source on ODA works well, there is in general an issue with performance of large joins or simple where clauses on large tables.
In my view, this is expected, on the source side the DB runs on EXA and uses the capabilities such as offloading and smart scan, but since the vDB is on ODA, these capabilities are missed and hence the slower performance.
Question : If the target host was also EXA, as in vDB would run on EXA, will the vDB still be able to use some of the features specific to EXA platform even though with the vDB we will not be using Oracle provided storage ?