Closed sticnarf closed 3 years ago
cc @5kbpers @MyonKeminta
We write the same lock when refreshes TTL too. I think CDC can handle it.
I write a new design of implementing fallback. It does not rewrite the primary lock (it's actually unnecessary, I think).
If we fallback from async commit using the solution 2 in https://github.com/tikv/sig-transaction/issues/64, we need to amend the primary lock clearing the async commit mark. Then, there can be two mutations writing the same lock. I am not sure whether CDC can handle this case.