support default primary key detection for mysql for int/temporal types

[jgao54]

When custom partition key is not specified, determine default partition key based on table's primary key. If the primary key(s) starts with an integer or temporal type, perform parallel snapshot instead of full table snapshot.

Testing:

• add unit & e2e tests
• validated parallel snapshotting locally

Fixes: DBI-867

[codecov]

❌ 2 Tests Failed:

Tests completed	Failed	Passed	Skipped
2304	2	2302	197

View the top 2 failed test(s) by shortest run time

github.com/PeerDB-io/peerdb/flow/e2e::TestApiPg

Stack Traces | 0.01s run time

=== RUN   TestApiPg
=== PAUSE TestApiPg
=== CONT  TestApiPg
--- FAIL: TestApiPg (0.01s)
2026/06/17 20:01:12 INFO Received AWS credentials from peer for connector: ci x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN}
2026/06/17 20:01:12 INFO Received AWS credentials from peer for connector: clickhouse x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN}
2026/06/17 20:01:12 INFO fetched schema x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} table=e2e_test_mychcl_64ugcxyz.test_time
2026/06/17 20:01:12 INFO fetched schema x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} table=e2e_test_mychcl_aldueesm.test_update_pkey_enabled

github.com/PeerDB-io/peerdb/flow/e2e::TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew

Stack Traces | 4.12s run time

=== RUN   TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew
=== PAUSE TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew
=== CONT  TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew
2026/06/17 19:59:36 INFO Received AWS credentials from peer for connector: ci x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN}
2026/06/17 19:59:36 INFO Received AWS credentials from peer for connector: clickhouse x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN}
2026/06/17 19:59:36 INFO Executing and processing query x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id"
2026/06/17 19:59:36 INFO Executing and processing query stream x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id"
2026/06/17 19:59:36 INFO [pg_query_executor] declared cursor x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart cursorQuery="DECLARE peerdb_cursor_5027000992335193366 CURSOR FOR SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id" args=[]
2026/06/17 19:59:36 INFO [pg_query_executor] fetching rows start x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id" channelLen=0
2026/06/17 19:59:36 INFO [pg_query_executor] fetching from cursor x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart cursor=peerdb_cursor_5027000992335193366
2026/06/17 19:59:36 INFO processed row stream x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart cursor=peerdb_cursor_5027000992335193366 records=2 bytes=19 channelLen=1
2026/06/17 19:59:36 INFO [pg_query_executor] fetched rows x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id" rows=2 bytes=19 channelLen=1
2026/06/17 19:59:36 INFO [pg_query_executor] fetching from cursor x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart cursor=peerdb_cursor_5027000992335193366
2026/06/17 19:59:36 INFO processed row stream x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart cursor=peerdb_cursor_5027000992335193366 records=0 bytes=0 channelLen=0
2026/06/17 19:59:36 INFO [pg_query_executor] fetched rows x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id" rows=0 bytes=0 channelLen=0
2026/06/17 19:59:36 INFO [pg_query_executor] committing transaction x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart
2026/06/17 19:59:36 INFO [pg_query_executor] committed transaction for query x-peerdb-additional-metadata={Operation:FLOW_OPERATION_UNKNOWN} partitionId=testpart query="SELECT id,val FROM e2e_test_api_k7lxhkpi.\"table1\" ORDER BY id" rows=2 bytes=19 channelLen=0
    api_test.go:3479: WaitFor wait for flow to be running 2026-06-17 19:59:40.258755513 +0000 UTC m=+405.352641579
    api_test.go:3490: 
        	Error Trace:	.../flow/e2e/api_test.go:3490
        	            				.../hostedtoolcache/go/1.26.4.../src/runtime/asm_amd64.s:1771
        	Error:      	"1" is not greater than "1"
        	Test:       	TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew
        	Messages:   	Should have multiple executions (continue-as-new happened)
    api_test.go:51: begin tearing down postgres schema api_azlkcktl
--- FAIL: TestApiPg/TestCreateCDCFlowAttachIdempotentAfterContinueAsNew (4.12s)

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[github-actions]

❌ Test Failure

Analysis: The PR's own new test Test_MySQL_Default_Partition_Key_Parallel_Snapshot deterministically times out during snapshot (STATUS_SNAPSHOT) in multiple matrix legs, indicating a real bug in the new MySQL default-partition-key parallel-snapshot feature rather than flakiness.
Confidence: 0.8

⚠️ This appears to be a real bug - manual intervention needed

View workflow run

[github-actions]

🔄 Flaky Test Detected

Analysis: The test failed on a non-retried assertion (require.Greater, "1 is not greater than 1") that races against Temporal's eventually-consistent ListWorkflow visibility indexing, and failed in only one of several parallel matrix jobs—both hallmarks of a flaky timing race rather than a real bug.
Confidence: 0.9

✅ Automatically retrying the workflow

View workflow run

[claude]

Code review

No issues found. Checked for bugs and CLAUDE.md compliance.

[ilidemi]

Any reason not to have all of this logic in the connector?

[jgao54]

mainly because it would require passing in the catalog pool to the connector method (needed by LoadTableSchemasFromCatalog); and in general it's cleaner separation of concern to have the connector focus on source db logic, while the generic activity handling the catalog logic. Note that there are already existing exceptions like PullRecords/PullQRepRecords, but trying not to introduce more of it.

The mysql-specific PeerDBMySQLDefaultPartitionKeyEnabled check being here is a bit awkward, but it will be removed once this feature stabilizes.

[ilidemi]

I think it's totally fine to introduce extra args that the connector is welcome to use or not use. The purpose of shared skeleton and interfaced connectors is that the skeleton doesn't have to have type switches. Very possible that better abstractions are out there but this one arg change doesn't seem like it's make things messy at all