[CRASH] Server got crashed while catching up one of the master - ASSERTION FAILED
keydb.log Crash report
KeyDB Version 6.3.2 (with flash enabled)
=== KEYDB BUG REPORT START: Cut & paste starting from here === 29967:30109:S 12 Jul 2023 16:53:57.328 # === ASSERTION FAILED === 29967:30109:S 12 Jul 2023 16:53:57.328 # ==> StorageCache.cpp:197 'bulkInsertsInProgress.load(std::memory_order_seq_cst) || count == (dictSize(m_pdict) + m_collisionCount)' is not true
------ STACK TRACE ------
Backtrace: ./keydb-server :6379(StorageCache::count() const+0x90) [0x5355b0] ./keydb-server :6379(redisDbPersistentData::size(bool) const+0x69) [0x5fa1f9] ./keydb-server :6379(redisDbPersistentData::ensure(char const, dictEntry)+0x2ca) [0x5fbd9a] ./keydb-server :6379() [0x5fc6a5] ./keydb-server :6379() [0x5e2b82] ./keydb-server :6379(rdbLoadRio(_rio, int, rdbSaveInfo)+0xa49) [0x5d77d9] ./keydb-server :6379(rdbLoadFile(char const, rdbSaveInfo, int)+0x4f) [0x5d95bf] ./keydb-server :6379(readSyncBulkPayloadRdb(connection, redisMaster*, rdbSaveInfo&, int&)+0x2ce) [0x5eecae] ./keydb-server :6379(readSyncBulkPayload(connection)+0xfc) [0x5f3e4c] ./keydb-server :6379(connSocketEventHandler(aeEventLoop, int, void*, int)+0x1a0) [0x548a80] ./keydb-server *:6379(ProcessEventCore+0x106) [0x63da56] ./keydb-server *:6379(aeProcessEvents+0x205) [0x63f035] ./keydb-server *:6379(aeMain+0x37) [0x63f647] ./keydb-server :6379(workerThreadMain(void)+0x77) [0x629e07] /lib64/libpthread.so.0(+0x7dd5) [0x7f6b73d78dd5] /lib64/libc.so.6(clone+0x6d) [0x7f6b73aa1ead]
Have attache the full backstrace and info with this thread
Aditional information
- OS distribution and version - CentOS Linux release 7.6.1810 (Core)
- Steps to reproduce (if any) - Have configured a multimaster replica with 3 master nodes. After startup, the server get crashed with error fastlock lock/unlock mismatch for: client 2677 #fastlock.cpp:295
@msotheeswaran-sc any news ?
There are quite alot of crashes in KeyDB that are unresolved @DaemonSnake