7 Pull requests merged by 6 people

• auth/passwords: Clean up prefix_for_scheme()

  #21749 merged Dec 4, 2024

• Use checksummed input streams in validate_checksums()

  #20933 merged Dec 4, 2024

• topology_coordinator: introduce reload_count in topology state and use it to prevent race

  #21713 merged Dec 3, 2024

• compaction: update maintenance sstable set on scrub compaction completion

  #21582 merged Dec 2, 2024

• tasks: add tablet repair virtual task

  #21624 merged Dec 2, 2024

• mutation: remove unused "#include"s

  #21727 merged Nov 29, 2024

• replica,sstables: track download progress of download_task_impl

  #21562 merged Nov 28, 2024

23 Pull requests opened by 16 people

• storage_proxy: set base_info for reverse view reads

  #21710 opened Nov 28, 2024

• truncate: commit log replay positions are not saved correctly

  #21722 opened Nov 28, 2024

• docs: add instruction for installing cassandra-stress

  #21723 opened Nov 28, 2024

• Reproduce the issue of non key attribute loss in GSI.

  #21725 opened Nov 29, 2024

• test: topology_custom: restrcuture comment as ordered list

  #21728 opened Nov 29, 2024

• tasks: add tablet migration virtual task

  #21729 opened Nov 29, 2024

• counters.hh: drop unused boost includes

  #21738 opened Dec 1, 2024

• .github: generate headers and use Debug build for header dependency verification

  #21739 opened Dec 2, 2024

• Materialized view name length should be limited

  #21746 opened Dec 2, 2024

• service/qos: increase timeout of internal get_service_levels queries

  #21748 opened Dec 2, 2024

• docs: Update documentation on CREATE ROLE WITH HASHED PASSWORD

  #21752 opened Dec 2, 2024

• install-dependencies.sh: cleanups to silence shellcheck

  #21756 opened Dec 3, 2024

• db/config: fix typo in description of repair_multishard_reader_enable_read_ahead

  #21758 opened Dec 3, 2024

• transport/server: revert using async function in `for_each_gently()`

  #21761 opened Dec 3, 2024

• sstables: parse(summary): reserve positions vector

  #21767 opened Dec 3, 2024

• dht: auto_refreshing_sharder.hh: don't include database.hh

  #21768 opened Dec 3, 2024

• build: cmake: Propagate target architecture to CMake build system

  #21771 opened Dec 4, 2024

• build: cmake: Improve function return mechanism with `return(PROPAGATE)`

  #21772 opened Dec 4, 2024

• service/qos/service_level_controller: update cache on startup

  #21773 opened Dec 4, 2024

• server: set `connection_stage` to ESTABLISHED when authenticated

  #21774 opened Dec 4, 2024

• Update mig-tool-review.rst

  #21775 opened Dec 4, 2024

• raft: clean up the documentation

  #21780 opened Dec 4, 2024

• table, storage_service: prepare and then apply replication map updates

  #21781 opened Dec 4, 2024

33 Issues closed by 15 people

• [aarch64, debug] topology_custom/test_data_resurrection_after_cleanup failed with

  #21548 closed Dec 4, 2024

• Bug in handling of shadowable tombstone and its affect on map

  #21769 closed Dec 4, 2024

• compaction: validate the checksums+digest of input sstables

  #19071 closed Dec 4, 2024

• [Epic] create native alternatives to java tools

  #14856 closed Dec 4, 2024

• Identify nodes by UUID for coordination and messaging instead of IPs

  #6403 closed Dec 4, 2024

• types: large allocations and stalls while comparing Decimal values

  #21716 closed Dec 3, 2024

• `storage_service::wait_for_topology_request_completion` may hang waiting for event after request was completed

  #19994 closed Dec 3, 2024

• Repair failed after RemoveNodeThenAddNode failed to properly remove node

  #10460 closed Dec 3, 2024

• Streaming after adding nodes causes to existing (not new) nodes run with 100% CPU on all shards

  #20921 closed Dec 3, 2024

• dtest cql_additional_tests :: test_nameless_index flaky with tablets enabled

  #21575 closed Dec 3, 2024

• use glob instead of os.join.path for figuring out non root nvmes

  #8778 closed Dec 3, 2024

• [Tablets] decommission failed: `lsa - Aborting due to allocation failure: failed to refill emergency reserve of 30 (have 29 free segments)`

  #18078 closed Dec 3, 2024

• doc: should explain task status retention and one-time query behavior

  #21757 closed Dec 3, 2024

• Recent conftest change broke pytest's default output format and made it too spammy

  #21712 closed Dec 2, 2024

• tasks-manager: create task for tablet repair

  #21368 closed Dec 2, 2024

• On Rocky8, nonroot install image causes "Unit scylla-server.service not found"

  #21720 closed Dec 2, 2024

• [6.1.0~rc1] nodetool scrub --skip-corrupted failed with filesystem error: link failed: No such file or directory

  #20030 closed Dec 2, 2024

• Optimize cleanup by being more precise about which sstables really need cleanup

  #828 closed Dec 2, 2024

• test bootstrap and replace of a zero-token node with inter-dc encryption enabled

  #21096 closed Dec 2, 2024

• [x86][dtest] TestBatch.test_replay_after_schema_change failed

  #17192 closed Dec 2, 2024

• test_misc.py::test_read_stats fails due to `There is no member named _raw.`

  #18700 closed Dec 2, 2024

• Make cqlpy as a python module

  #21644 closed Dec 2, 2024

• test_raft_recovery_stuck.test_recover_stuck_raft_recovery fails due to creating table in non-existent keyspace

  #21724 closed Nov 29, 2024

• [x86_64, release] object_store/test_backup failed with

  #20757 closed Nov 29, 2024

• ALTER KEYSPACE 'keyspace_new_dc' Fails with "Keyspace Doesn't Exist" Error

  #21337 closed Nov 29, 2024

• Compacting of sstables interrupted due to `sstables::malformed_sstable_exception`: Checksummed chunk of size 0 at file offset

  #21650 closed Nov 29, 2024

• Dump diagnostics on SIGQUIT

  #7400 closed Nov 28, 2024

• scylla-sstable: remove --scylla-yaml-file as schema-source

  #20534 closed Nov 28, 2024

• Initial-tablets is not propagated properly via topology-coordinator

  #20768 closed Nov 28, 2024

• Removenode / Decommission not working

  #21705 closed Nov 28, 2024

• hinted handoff: Possible dereferencing a null pointer

  #21699 closed Nov 28, 2024

• Monitor restore process

  #21427 closed Nov 28, 2024

• docs: Issue on page Nodetool tablestats

  #21586 closed Nov 28, 2024

23 Issues opened by 18 people

• Balancer is not fairly distributing tablets in a heterogeneous cluster during decommission

  #21783 opened Dec 5, 2024

• [x86_64, release] boost/secondary_index_test failed with Entering

  #21782 opened Dec 4, 2024

• DDL failing with "Failed to apply group 0 change due to concurrent modification"

  #21779 opened Dec 4, 2024

• Could not create SSTable component due to filesystem_error (filesystem error: open failed: No such file or directory)

  #21778 opened Dec 4, 2024

• Converts the rest of the system to identify nodes by host ids

  #21777 opened Dec 4, 2024

• streaming/repair: validate checksum+digest of sstables

  #21776 opened Dec 4, 2024

• [x86_64, release] boost/logalloc_test failed with Leaving

  #21770 opened Dec 3, 2024

• [Perf][Tablets] Latency spike up to 86 ms during steady step, when no cluster operations running

  #21766 opened Dec 3, 2024

• Node remove failed during disrupt_remove_node_then_add_node

  #21765 opened Dec 3, 2024

• service level cache may not be populated at startup

  #21763 opened Dec 3, 2024

• Finalize tablet splits earlier

  #21762 opened Dec 3, 2024

• Enable enable_small_table_optimization_for_rbno by default

  #21760 opened Dec 3, 2024

• main, sstables: add summary of sstables in-memory footprint to diagnostics dumped on SIGQUIT

  #21751 opened Dec 2, 2024

• update_cluster_layout_tests.TestUpdateClusterLayout.test_simple_add_new_node_while_query_info fails with Client request timeout

  #21750 opened Dec 2, 2024

• Scylla nodes restarted due to no HOST_ID found in application_state

  #21744 opened Dec 2, 2024

• test.py should consider a test fail if scylla coredumps or crashes

  #21742 opened Dec 2, 2024

• tablets merge is not triggered after 80% capacity shrink

  #21736 opened Dec 1, 2024

• raft + topology-on-raft: dump diagnostics on SIGQUIT

  #21731 opened Nov 29, 2024

• improvement: compress compaction logs when compacting files in common dir

  #21730 opened Nov 29, 2024

• truncate: commit log replay positions are not saved correctly

  #21719 opened Nov 28, 2024

• raft topology: failed replace-with-same-IP leaves wrong host ID in gossiper, preventing retry until contact point is restarted

  #21718 opened Nov 28, 2024

• Node crashes when running a test with read repair and trace log level

  #21714 opened Nov 28, 2024

• Revert scylladb/scylladb#21658 when scylladb/scylladb#21534 will be fixed

  #21711 opened Nov 28, 2024

102 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Fix Replication Factor calculation for Tablets effective replication map.

  #21289 commented on Dec 4, 2024 • 26 new comments

• Rack aware view pairing

  #21453 commented on Dec 4, 2024 • 22 new comments

• Add tablet merge support

  #20572 commented on Dec 5, 2024 • 20 new comments

• Alternator add WCU for delelte item

  #21709 commented on Dec 4, 2024 • 16 new comments

• Fix gossiper orphan node floating problem by adding a remover fiber

  #21600 commented on Dec 4, 2024 • 14 new comments

• truncate: make TRUNCATE TABLE safe with tablets

  #19789 commented on Dec 4, 2024 • 13 new comments

• mv: handle mismatched base/view replica count caused by RF change

  #21651 commented on Dec 4, 2024 • 9 new comments

• service/topology_coordinator: migrate view builder only if all nodes are up

  #21708 commented on Dec 4, 2024 • 8 new comments

• repair: wait for view update to complete

  #21657 commented on Dec 4, 2024 • 5 new comments

• Reserve IOCBs for tool applications

  #21527 commented on Dec 4, 2024 • 4 new comments

• Enhance load-and-stream with "scope"

  #21169 commented on Dec 5, 2024 • 2 new comments

• materialized view: make flow-control maximum delay configurable

  #21647 commented on Dec 4, 2024 • 2 new comments

• Atomic in-memory schema changes application

  #20853 commented on Dec 4, 2024 • 2 new comments

• utils: add disk_space_monitor

  #21523 commented on Dec 2, 2024 • 2 new comments

• cql: Add support for IN clause on indexed column without filtering

  #21683 commented on Nov 28, 2024 • 2 new comments

• test.py: Add the possibility to run boost and unit tests with pytest

  #21108 commented on Dec 3, 2024 • 0 new comments

• [Backport 6.1] sstables: Fix use-after-free on page cache buffer when parsing promoted index entries across pages

  #20906 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.2] utils: cached_file: Mark permit as awaiting on page miss

  #21358 commented on Dec 2, 2024 • 0 new comments

• docs: rest.rst: use latest docker tag to view Swagger UI for REST API

  #21681 commented on Dec 4, 2024 • 0 new comments

• repair: Support repair_history migration after topology change

  #18211 commented on Dec 4, 2024 • 0 new comments

• locator::topology: add structured datacenter and rack classes

  #17273 commented on Dec 1, 2024 • 0 new comments

• `[Service = cdc::generation_service]: Assertion 'local_is_initialized()' failed` during shutdown after a failed boot

  #21534 commented on Dec 5, 2024 • 0 new comments

• Large (~850ms) stalls on 5000 tables tests

  #21646 commented on Dec 4, 2024 • 0 new comments

• [tablets] Decommission failure: `Failed to drain tablets: std::runtime_error (Unable to find new replica for tablet)`

  #19504 commented on Dec 4, 2024 • 0 new comments

• Materialized view name length should be limited

  #20755 commented on Dec 4, 2024 • 0 new comments

• [C-s] failed during few nemises run in parallel (disrupt_create_index, disrupt_decommission_streaming_err) in longevity multi dc test (12h)

  #21120 commented on Dec 4, 2024 • 0 new comments

• During repair and drop index next error appeared: Unexpected exception while revoking all permissions on dropped table service::group0_concurrent_modification (Failed to apply group 0 change due to concurrent modification)

  #20560 commented on Dec 4, 2024 • 0 new comments

• Node deletion while create index operation led to index is never built

  #21695 commented on Dec 4, 2024 • 0 new comments

• scrub/segregate: add a flag which makes it drop all sstables it cannot fix

  #19060 commented on Dec 4, 2024 • 0 new comments

• scrub: handle malformed_sstable_exception

  #19059 commented on Dec 4, 2024 • 0 new comments

• scrub/validate: check digest of all sstable components

  #20103 commented on Dec 4, 2024 • 0 new comments

• test: Move test_query_built_indexes_virtual_table from boost to cqlpy

  #21677 commented on Dec 3, 2024 • 0 new comments

• Improve load shedding on the replica side

  #21649 commented on Dec 2, 2024 • 0 new comments

• [Backport 6.1] test/boost/view_schema_test.cc: Wait for views to build in test_view_update_generating_writetime

  #21639 commented on Dec 5, 2024 • 0 new comments

• [Backport 6.2] docs: correct the storage size for n2-highmem-32 to 9000GB

  #21595 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.1] docs: correct the storage size for n2-highmem-32 to 9000GB

  #21594 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.2] docs: fix monospace formatting for `rm` command

  #21593 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.1] docs: fix monospace formatting for `rm` command

  #21592 commented on Dec 4, 2024 • 0 new comments

• test: Add validation of getting/changing compaction strategy via REST…

  #21569 commented on Dec 3, 2024 • 0 new comments

• Make restore task abortable

  #21567 commented on Dec 2, 2024 • 0 new comments

• Add tests for MVs and indexes representation in system. table(s) and API(s)

  #21551 commented on Nov 28, 2024 • 0 new comments

• [Backport 6.2] dist: systemd: use default KillMode

  #21510 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.1] dist: systemd: use default KillMode

  #21509 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.2] test: rename "cql-pytest" to "cqlpy"

  #21488 commented on Dec 2, 2024 • 0 new comments

• [Backport 6.1] test: rename "cql-pytest" to "cqlpy"

  #21484 commented on Dec 2, 2024 • 0 new comments

• test/topology_experimental_raft: Prevent keyspace creation conflicts with IF NOT EXISTS

  #21702 commented on Dec 2, 2024 • 0 new comments

• test: compile all unit tests into a single executable

  #21443 commented on Nov 28, 2024 • 0 new comments

• locator: topology: use const node& instead of const node*

  #21439 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.1] test/topology_experimental_raft: fix flaky test

  #21396 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.2] test/topology_experimental_raft: fix flaky test

  #21395 commented on Dec 4, 2024 • 0 new comments

• [Backport 6.1] utils: cached_file: Mark permit as awaiting on page miss

  #21359 commented on Dec 4, 2024 • 0 new comments

• sstable: add digest for all sstable components

  #20100 commented on Dec 4, 2024 • 0 new comments

• DESCRIBE SCHEMA for alternator table in restore context

  #19112 commented on Dec 2, 2024 • 0 new comments

• Help drivers to invalidate tablets in time

  #21664 commented on Dec 2, 2024 • 0 new comments

• test_crash_coordinator_before_streaming.test_kill_coordinator_during_op.debug.1 times out

  #21114 commented on Dec 2, 2024 • 0 new comments

• raft topology / tablets: introduce nodetool API to mark node as permanently dead (like in `removenode` but without the actual remove)

  #21281 commented on Dec 1, 2024 • 0 new comments

• Support fast sstable cleanup

  #12385 commented on Dec 1, 2024 • 0 new comments

• replica-side load shedding is not aggressive enough

  #14909 commented on Dec 1, 2024 • 0 new comments

• Can't drop/alter types in keyspace

  #5155 commented on Dec 1, 2024 • 0 new comments

• doc: improve the documentation for tablets

  #19209 commented on Dec 1, 2024 • 0 new comments

• Alternator: performance idea: cache parsed expressions

  #5023 commented on Nov 29, 2024 • 0 new comments

• [CI] test_tablets.test_read_of_pending_replica_during_migration[true] flaky

  #20844 commented on Nov 29, 2024 • 0 new comments

• [2024.1] TestUpdateClusterLayout.test_simple_kill_node_while_decommissioning failed in OperationTimedOut

  #19352 commented on Nov 29, 2024 • 0 new comments

• Node startup failed during rebootstrap process - raft_operation_timeout_error (group [693de0d0-a3d1-11ef-a229-81247977f4b8] raft operation [read_barrier] timed out)

  #21611 commented on Nov 29, 2024 • 0 new comments

• Internal reads from group 0 tables are using the wrong semaphore, causing group 0 unavailability due to getting stuck behind user generated workload

  #21637 commented on Nov 29, 2024 • 0 new comments

• Node startup failed during bootstrap process - raft_operation_timeout_error (group [b4705300-a951-11ef-95b1-db8753ca301d] raft operation [read_barrier] timed out)

  #21674 commented on Nov 29, 2024 • 0 new comments

• tasks-manager: create tasks for tablet migration

  #21365 commented on Nov 29, 2024 • 0 new comments

• test_topology_recovery_basic fails when validation queries are done with all_pages=True

  #19101 commented on Nov 29, 2024 • 0 new comments

• Major bug: Repair will cause index table data loss!

  #20808 commented on Nov 29, 2024 • 0 new comments

• [Raft Topology] 'OperationTimedOut' error in SELECT query during adding new node

  #17445 commented on Nov 29, 2024 • 0 new comments

• raft topology: (design) external recovery tool for majority loss and group 0 bugs

  #18791 commented on Nov 28, 2024 • 0 new comments

• raft topology: (implementation) external recovery tool for majority loss and group 0 bugs

  #20657 commented on Nov 28, 2024 • 0 new comments

• gossip topology: Failure during replace-with-same-IP leaves the node without `STATUS` application_state (permanently), and `token_metadata` inconsistent (until restart) (applies to gossiper / "node-ops" based topology changes)

  #19975 commented on Nov 28, 2024 • 0 new comments

• Integrated Backup and Restore in Scylla

  #21100 commented on Nov 28, 2024 • 0 new comments

• A lot of seastar::named_semaphore_timed_out (Semaphore timed out: _read_concurrency_sem) cause gemini to failed

  #10210 commented on Nov 28, 2024 • 0 new comments

• doc: link to SBOM from the docs

  #18195 commented on Nov 28, 2024 • 0 new comments

• Tried to build a global schema for view keyspace.materializedView with an uninitialized base info

  #21354 commented on Nov 28, 2024 • 0 new comments

• docs: add comprehensive documentation on scrub

  #19062 commented on Dec 4, 2024 • 0 new comments

• REST API: add endpoint which drops all quarantined sstables

  #19061 commented on Dec 4, 2024 • 0 new comments

• [Epic] Scrub as a reliable means to clean nodes of bad sstables

  #15693 commented on Dec 4, 2024 • 0 new comments

• system.clients shows connections as AUTHENTICATING, not READY, even after they pass authentication

  #12640 commented on Dec 4, 2024 • 0 new comments

• repair_history data not migrated along with tablet

  #17507 commented on Dec 4, 2024 • 0 new comments

• Coordinator node timed out waiting for replica nodes' responses while read `system.scylla_local` table from new added node ( during read load scenario only)

  #20235 commented on Dec 3, 2024 • 0 new comments

• [x86_64, debug] topology_random_failures/test_random_failures failed with Operation timed out for system.service_levels_v2

  #20483 commented on Dec 3, 2024 • 0 new comments

• dtest, test_raft_recovery_procedure: new CDC generation from raft topology upgrade procedure not handled by every node; instead, legacy CDC code tries to obtain it (and fails)

  #21227 commented on Dec 3, 2024 • 0 new comments

• Scylla hangs on shutdown

  #16645 commented on Dec 3, 2024 • 0 new comments

• Do not logs on large partitions - that are less than 1000MB

  #20125 commented on Dec 3, 2024 • 0 new comments

• Tablets split never concludes in the presence of heavy write workload

  #20054 commented on Dec 3, 2024 • 0 new comments

• Cluster went into invalid state: "Host 10.0.9.3 does not have HOST_ID application_state" + "Raft server id cannot be translated to an IP address" after clean restart

  #21666 commented on Dec 3, 2024 • 0 new comments

• Create schema with autoresize rf for multidc cluster with tablets enabled and configured with on dc with single zero token node failed and terminate the c-s

  #20684 commented on Dec 3, 2024 • 0 new comments

• test_schema_and_data_on_mixed_versions_cluster failed with 'Unable to complete the operation against any hosts'

  #21704 commented on Dec 3, 2024 • 0 new comments

• (tablets only) All DC nodes got Coredump and error of: "failed to log message: fmt='Requested location for node {} not in topology" on altering keyspace replication-factor of one data-center (out of 2)

  #20282 commented on Dec 3, 2024 • 0 new comments

• [Tablets] Degraded max throughput in comparison to vnodes

  #19108 commented on Dec 3, 2024 • 0 new comments

• Cannot use default system-trust with deb-packaged scylla

  #13904 commented on Dec 3, 2024 • 0 new comments

• alternator_tests.test_slow_query_logging failed in configuration_exception (Unrecognized strategy option {UNKNOWN_DC}..)

  #21490 commented on Dec 2, 2024 • 0 new comments

• authentication: Document CREATE ROLE WITH HASHED PASSWORD as Cassandra's equivalent

  #21691 commented on Dec 2, 2024 • 0 new comments

• Sporadic SIGSEGV (exit code -11) in Python when exiting the Alternator test

  #17564 commented on Dec 2, 2024 • 0 new comments

• Make the commit log hard limit mandatory

  #16471 commented on Dec 2, 2024 • 0 new comments

• 55 partitions are (temporarily) missing in a select query of scylla-bench table

  #20084 commented on Dec 2, 2024 • 0 new comments

• test_raft_service_levels.py::test_service_levels_upgrade fails due to "signed integer overflow"

  #19348 commented on Dec 2, 2024 • 0 new comments

• toJson() truncates some doubles to integers

  #7972 commented on Dec 2, 2024 • 0 new comments

• Tablet RESTful APIs which transition a tablet executed during decommission will put the state machine out of the decommission track

  #20073 commented on Dec 2, 2024 • 0 new comments