-
Type: Build Failure
-
Resolution: Duplicate
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
5
-
Storage - Ra 2020-06-15, Storage - Ra 2020-06-29
A multiversion test suite in MongoDB is failing during rollback to stable, with an assertion indicating that timestamps aren't in an expected state.
The test case that triggered the failure is auto-generated in MongoDB, so I can't link to a particular test case.
The first failure happened after this drop from WiredTiger - so a good starting point for diagnosing is probably reviewing the changes included there.
The assertion is:
[js_test:rollback_test-536d2-1591173474752-1] 2020-06-03T09:00:44.654+0000 d20031| | 2020-06-03T09:00:44.654+00:00 E STORAGE 22435 [SignalHandler] "WiredTiger error","attr":{"error":0,"message":"[1591174844:654467][45275:0x7f47a189b700], file:_mdb_catalog.wt, txn rollback_to_stable: __rollback_row_ondisk_fixup_key, 246: (newer_hs_ts == WT_TS_NONE || hs_stop_ts <= newer_hs_ts || hs_start_ts == hs_stop_ts)"}
Stacktrace:
mongo::stack_trace_detail::(anonymous namespace)::printStackTraceImpl(mongo::stack_trace_detail::(anonymous namespace)::Options const&, mongo::StackTraceSink*) mongo::printStackTrace() mongo::(anonymous namespace)::abruptQuit(int) funlockfile gsignal abort __wt_abort __rollback_row_ondisk_fixup_key __rollback_abort_row_reconciled_page_internal __rollback_abort_newer_updates __rollback_to_stable_btree_apply __wt_rollback_to_stable __wt_txn_global_shutdown __conn_close mongo::WiredTigerKVEngine::cleanShutdown() mongo::shutdownGlobalStorageEngineCleanly(mongo::ServiceContext*) mongo::(anonymous namespace)::shutdownTask(mongo::ShutdownTaskArgs const&) mongo::(anonymous namespace)::runTasks(std::stack<mongo::unique_function<void (mongo::ShutdownTaskArgs const&)>, std::deque<mongo::unique_function<void (mongo::ShutdownTaskArgs const&)>, std::allocator<mongo::unique_function<void (mongo::ShutdownTaskArgs const&)> > > >, mongo::ShutdownTaskArgs const&) mongo::shutdown(mongo::ExitCode, mongo::ShutdownTaskArgs const&) mongo::exitCleanly(mongo::ExitCode) mongo::(anonymous namespace)::handleOneSignal(mongo::(anonymous namespace)::SignalWaitResult const&, mongo::(anonymous namespace)::LogRotationState*) mongo::(anonymous namespace)::signalProcessingThread(mongo::LogFileStatus) mongo::stdx::thread::thread<void (*)(mongo::LogFileStatus), mongo::LogFileStatus&, 0>(void (*)(mongo::LogFileStatus), mongo::LogFileStatus&)::{lambda()#1}::operator()() execute_native_thread_routine start_thread clone
- depends on
-
WT-6388 Fix-up out-of-order updates in the history store
- Closed
- related to
-
SERVER-48772 Dropping multiple indexes incorrectly timestamps writes
- Closed