Doing usual rebase is complicated as the files have been renamed/modified on both branches. I think it's probably easier to just create another branch on top of master and try to move stuff that works from old branch without keeping whole history.

Here is  the summary of changes that were applied to u/andy-slac/cassandra branch since commit c7c8782:

1. (406935c) renamed ppdbSchema.py to ppdbSqlSchema.py (no changes to the code)
2. (01e5696) factrored out PpdbBaseSchemaConfig and PpdbBaseSchema from the above, these are base classes with a functionality common to all implementations.
   • new class PpdbSqlSchemaConfig inheriting from PpdbBaseSchemaConfig
   • PpdbConfig now inherits from PpdbSqlSchemaConfig
   • PPdbSchema renamed to PpdbSqlSchema inheriting PpdbBaseSchema
   • small changes to Ppdb class
   • test_ppdbSchema.py updated
3. (ced16f0) implemented PpdbCassandraSchemaConfig(PpdbBaseSchemaConfig), PpdbCassandraSchema(PpdbBaseSchema), PpdbCassandraConfig(PpdbCassandraSchemaConfig) and no-op PpdbCassandra class
4. (ea1dd4d) added implementation for few PpdbCassandra methods
5. (1538a8e) added couple of more store methods
6. (4e65d1d) added getDiaObjects implementation
7. (b2349b9) implemented getDiaSources
8. (53145bd4) improved query generation
9. (a014b64c) month-based partitioning for sources
10. (c656916) change hard-coded timeout value

The non-trivial part here is refactoring of sql-based implementation (first two commits), I'm going to repeat that process to make sure that nothing breaks and then just copy Cassandra implemention in its current state from old branch (and rename ppdb -> apdb)

Similarly I kept old branch u/andy-slac/cassandra in l1dbproto and replayed it on top of the new master making new branch u/andy-slac/cassandra-2.

Merged all changes, u/andy-slac/cassandra-2 is a new development branch for both packages. Tested that it works OK on my dev VM.