Hi,
I'm trying to run a UNION ALL query on two LookupTableSource tables defined
with JDBC. When attempting this Flink complains that this is an unsupported
feature:
Caused by: org.apache.calcite.plan.RelOptPlanner$CannotPlanException: There
are not enough rules to produce a node with desired properties:
convention=STREAM_PHYSICAL, FlinkRelDistributionTraitDef=any,
MiniBatchIntervalTraitDef=None: 0, ModifyKindSetTraitDef=[NONE],
UpdateKindTraitDef=[NONE].
Missing conversions are FlinkLogicalTableSourceScan[convention: LOGICAL ->
STREAM_PHYSICAL] (2 cases)
There are 2 empty subsets:
Empty subset 0: rel#191:RelSubset#4.STREAM_PHYSICAL.any.None:
0.[NONE].[NONE], the relevant part of the original plan is as follows
168:FlinkLogicalTableSourceScan(table=[[default_catalog, default_database,
table]], fields=[...])
Empty subset 1: rel#192:RelSubset#5.STREAM_PHYSICAL.any.None:
0.[NONE].[NONE], the relevant part of the original plan is as follows
170:FlinkLogicalTableSourceScan(table=[[default_catalog, default_database,
other_table]], fields=[...])
Root: rel#189:RelSubset#7.STREAM_PHYSICAL.any.None: 0.[NONE].[NONE]
Original rel:
FlinkLogicalSink(subset=[rel#166:RelSubset#3.LOGICAL.any.None:
0.[NONE].[NONE]], table=[default_catalog.default_database.table_output],
fields=[...]): rowcount = 2.0E8, cumulative cost = {2.0E8 rows, 2.0E8 cpu,
0.0 io, 0.0 network, 0.0 memory}, id = 174
FlinkLogicalUnion(subset=[rel#173:RelSubset#2.LOGICAL.any.None:
0.[NONE].[NONE]], all=[true]): rowcount = 2.0E8, cumulative cost = {2.0E8
rows, 0.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 172
FlinkLogicalTableSourceScan(subset=[rel#169:RelSubset#0.LOGICAL.any.None:
0.[NONE].[NONE]], table=[[default_catalog, default_database, table]],
fields=[...]): rowcount = 1.0E8, cumulative cost = {1.0E8 rows, 1.0E8 cpu,
1.32E10 io, 0.0 network, 0.0 memory}, id = 168
FlinkLogicalTableSourceScan(subset=[rel#171:RelSubset#1.LOGICAL.any.None:
0.[NONE].[NONE]], table=[[default_catalog, default_database, other_table]],
fields=[...]): rowcount = 1.0E8, cumulative cost = {1.0E8 rows, 1.0E8 cpu,
1.32E10 io, 0.0 network, 0.0 memory}, id = 170
I do understand that the semantics of unioning two lookups may be a bit
complicated, but was wondering if this is planned to be supported in the
future?
--
Best Regards,
Yuval Itzchakov.
