问题现象
OBServer 随机 core，查看堆栈发现内存状态正常。crash 信息为 sig=7,sig_code=4。
关键信息
通过 grep 命令过滤出 observer.log 日志中有关 CRASH 的信息，能看到 sig=7,sig_code=4。
查看 dmesg 信息，能够看到有硬件故障报错，日志信息如下。
[13739087.142412] mce: [Hardware Error]: Machine check events logged
[13739087.190098] Disabling lock debugging due to kernel taint
[13739087.190103] mce: [Hardware Error]: Machine check events logged
[13739087.210816] EDAC skx MC7: HANDLING MCE MEMORY ERROR
[13739087.210820] EDAC skx MC7: CPU 32: Machine Check Event: 0x0 Bank 25: 0xac00000200a00090
[13739087.210822] mce: Uncorrected hardware memory error in user-access at 60ae6eaa80
[13739087.219562] MCE 0x60ae6ea: Killing observer:18651 due to hardware memory corruption
[13739087.219566] MCE 0x60ae6ea: dirty LRU page recovery: Recovered
[13739087.235839] EDAC skx MC7: TSC 0x0
[13739087.235841] EDAC skx MC7: ADDR 0x60ae6eaa80
[13739087.235842] EDAC skx MC7: MISC 0x80008080b994c86
[13739087.235842] EDAC skx MC7: PROCESSOR 0:0x606a6 TIME 1751438335 SOCKET 1 APIC 0x80
[13739087.235850] EDAC MC7: 1 UE memory read error on CPU_SrcID#1_MC#3_Chan#0_DIMM#0 (channel:0 slot:0 page:0x60ae6ea offset:0xa80 grain:32 - err_code:0x00a0:0x0090 SystemAddress:0x60ae6eaa80 ProcessorSocketId:0x1 MemoryControllerId:0x3 ChannelAddress:0x405cdd480 ChannelId:0x0 RankAddress:0x202e6f480 PhysicalRankId:0x0 DimmSlotId:0x0 Row:0x10173 Column:0x298 Bank:0x2 BankGroup:0x0 ChipSelect:0x0 ChipId:0x0)
[13739173.848383] mce: [Hardware Error]: Machine check events logged
[13739173.855908] EDAC skx MC7: HANDLING MCE MEMORY ERROR
[13739173.855910] EDAC skx MC7: CPU 32: Machine Check Event: 0x0 Bank 25: 0x8c00008200800090
[13739173.855911] EDAC skx MC7: TSC 0x0
[13739173.855913] EDAC skx MC7: ADDR 0x60ae6c3400
[13739173.855913] EDAC skx MC7: MISC 0x80008080b987086
[13739173.855914] EDAC skx MC7: PROCESSOR 0:0x606a6 TIME 1751438422 SOCKET 1 APIC 0x80
[13739173.855920] EDAC MC7: 2 CE memory read error on CPU_SrcID#1_MC#3_Chan#0_DIMM#0 (channel:0 slot:0 page:0x60ae6c3 offset:0x400 grain:32 syndrome:0x0 - err_code:0x0080:0x0090 SystemAddress:0x60ae6c3400 ProcessorSocketId:0x1 MemoryControllerId:0x3 ChannelAddress:0x405cd8700 ChannelId:0x0 RankAddress:0x202e6c700 PhysicalRankId:0x0 DimmSlotId:0x0 Row:0x10173 Column:0xe0 Bank:0x2 BankGroup:0x0 ChipSelect:0x0 ChipId:0x0)
问题原因
机器出硬件故障，导致 OBServer core 掉。
问题的风险及影响
OBServer core 掉。
适用版本
OceanBase 数据库所有版本。
解决方法及规避方式
硬件故障，无法解决。只能 core 掉之后重启，切换机器。
