Denis Rystsov

Subj

Consumer failed to consume up to offsets in `SIPartitionMovementTest.test_shadow_indexing`

11

https://buildkite.com/redpanda/redpanda/builds/9985#74311a35-dfef-42fb-a669-5e752e0248a0 ``` Module: rptest.tests.partition_movement_test Class: SIPartitionMovementTest Method: test_shadow_indexing ``` ``` test_id: rptest.tests.partition_movement_test.SIPartitionMovementTest.test_shadow_indexing status: FAIL run time: 2 minutes 1.397 seconds TimeoutError("Consumer failed to consume up to offsets {TopicPartition(topic='topic', partition=0): 3271}...

Transactions: remove expired abort indexes

1

## Cover letter Once a log was evicted up to an offset X it's wasteful to store abort indexes covering the offsets before X (the abort index is used only...

Reduce duration of partition_movement_test from 25min to 8min

5

## Cover letter - reduces duration of partition_movement_test from 25min to 8min - introduce a workload generator / consistency checker to verified producer / consumers The new verifier is based...

Producer gets stuck in init_producer_id loop

2

Description =========== Observed a pathological behavior which is missing in other clients (java's `kafka-clients` & `franz-go`). With disabled transactions when any client (including `confluent_kafka`) is about to send `init_producer_id` it...

CI Failure (heap-use-after-free) in TxFeatureFlagTest.test_disabling_transactions_after_they_being_used

5

https://buildkite.com/redpanda/redpanda/builds/15094#018317d5-4e4a-4a52-8bad-9de3f437e31b ``` Module: rptest.tests.tx_feature_flag_test Class: TxFeatureFlagTest Method: test_disabling_transactions_after_they_being_used ``` ``` test_id: rptest.tests.tx_feature_flag_test.TxFeatureFlagTest.test_disabling_transactions_after_they_being_used status: FAIL run time: 2 minutes 40.395 seconds Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/ducktape/tests/runner_client.py", line 135, in...

CI Failure in UpgradeBackToBackTest.test_upgrade_with_all_workloads

10

https://buildkite.com/redpanda/redpanda/builds/15263#0183381a-ce92-4a7a-9d5f-da89c6dbdbab ``` Module: rptest.tests.upgrade_test Class: UpgradeBackToBackTest Method: test_upgrade_with_all_workloads Arguments: { "single_upgrade": false } ``` ``` test_id: rptest.tests.upgrade_test.UpgradeBackToBackTest.test_upgrade_with_all_workloads.single_upgrade=False status: FAIL run time: 13 minutes 23.172 seconds TimeoutError('') Traceback (most recent call...

Fix test_restore_with_aborted_tx

1

## Cover letter PR adds retries to give a transactional coordinator to elect a leader after a cluster is started to prevent NOT_COORDINATOR. Fixes #6328 ## Backport Required - [x]...

txn coordinator: gate tm_stm for long txn operations

1

## Cover letter It should fix a use-after-free problem caused by `tm_stm` shutdown interweaving with end_txn Fixes #6329 ## Backport Required - [ ] not a bug fix - [...

Review each unknown_server_error in the transactinal area (tx_gateway_frontend, rm_stm)

2

`unknown_server_error` is a fatal error (an app is required to recreate a producer) and we should return it only if it's the only way to handle the situation.