* Use serviceerror.Unavailable error type instead of serviceerror.Internal
* serviceerror.Unavailable indicates errors like service's dependency is down, retryable
* serviceerror.Internal indicates errors like NPE in service logic, non-retryable
* SQL: Rename SelectFromHistoryNode to RangeSelectFromHistoryNode
* SQL: Rename DeleteFromHistoryNode to RangeDeleteFromHistoryNode
* SQL: New DeleteFromHistoryNode aiming to delete one record from history node table
* New InternalHistoryNode struct containing history node info
* Modify SQL persistence layer to use key set pagination, instead of existing hacky solution
* ReadHistoryBranch in addition will allow returning history node metadata only
* Move history node filtering logic from persistence layer to upper history store layer
* New last_history_node_txn_id within execution info tracking the last valid history node
* New prev_txn_id within history node table tracking previous valid history node record
* Add db version attribute to executions table for better CAS guarantee
* 1.9.x will by default use existing next event ID for CAS, but will also support the new DB version CAS
* 1.10.x will by default use DB version CAS
Last event ID is not enough for CAS operation protection: e.g. activity heartbeat / buffer events.
The newly introduced DB version will be increased by 1 each time execution table is updated (update to mutable state record).
* Try using a set of default database names when initializing admin connection
Aiven DB is a PostgreSQL compatible DB, but comes with a different default DB name
* Add version check method to SQL interface
* Remove tools version check method in favor of functions defined in persistence package
* Refactor tests to not depend on tools package
* Remove forced namespace cache refresh
* Make SQL insert into history tree idempotent
NOTE: Previously, if a namespace is newly created and not loaded
in cache, the namespace cache can do forced cache refresh reducing
the wait time for caller. This forced cache refreshment may able
to trigger a new callback for failover. Now, if a task with deleted
namespace is processed, namespace cache will trigger the above
cache refreshment and causing a deadlock. The only solution now is
to remove the namespace cache forced refreshment logic.
* Optimize SQL layer supporting batch delete
* Split DeleteFromActivityInfoMaps into DeleteFromActivityInfoMaps and DeleteAllFromActivityInfoMaps
* Split DeleteFromTimerInfoMaps into DeleteFromTimerInfoMaps and DeleteAllFromTimerInfoMaps
* Split DeleteFromChildExecutionInfoMaps into DeleteFromChildExecutionInfoMaps and DeleteAllFromChildExecutionInfoMaps
* Split DeleteFromRequestCancelInfoMaps into DeleteFromRequestCancelInfoMaps and DeleteAllFromRequestCancelInfoMaps
* Split DeleteFromSignalInfoMaps into DeleteFromSignalInfoMaps and DeleteAllFromSignalInfoMaps
* Split DeleteFromSignalsRequestedSets into DeleteFromSignalsRequestedSets and DeleteAllFromSignalsRequestedSets
* Split ActivityInfoMapsFilter into ActivityInfoMapsSelectFilter and ActivityInfoMapsDeleteFilter
* Split TimerInfoMapsFilter into TimerInfoMapsSelectFilter and TimerInfoMapsDeleteFilter
* Split ChildExecutionInfoMapsFilter into ChildExecutionInfoMapsSelectFilter and ChildExecutionInfoMapsDeleteFilter
* Split RequestCancelInfoMapsFilter into RequestCancelInfoMapsSelectFilter and RequestCancelInfoMapsDeleteFilter
* Split SignalInfoMapsFilter into SignalInfoMapsSelectFilter and SignalInfoMapsDeleteFilter
* Split SignalsRequestedSetsFilter into SignalsRequestedSetsSelectFilter and SignalsRequestedSetsDeleteFilter
* When performing workflow reset & terminating current workflow, one unnecessary create history tree API is called, this PR remove the redundant call to improve performance
