Possíveis explicações de porque meu servidor ficou sem resposta

Question

Pouco antes de meu servidor não responder por alguns minutos, encontrei as seguintes linhas de log que parecem relacionadas. Espero entender melhor o que eles significam e sob quais condições eles aconteceriam:

Aug 25 18:23:32 myserver journal: Runtime journal is using 776.0M (max allowed 793.9M, trying to leave 1.1G free of 6.9G available → current limit 793.9M).
Aug 25 18:23:32 myserver journal: Runtime journal is using 776.0M (max allowed 793.9M, trying to leave 1.1G free of 6.9G available → current limit 793.9M).
Aug 25 18:23:32 myserver kernel: INFO: task in:imjournal:2125 blocked for more than 120 seconds.
Aug 25 18:23:32 myserver kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Aug 25 18:23:32 myserver kernel: in:imjournal    D ffff88042bd2b8c8     0  2125      1 0x00000080
Aug 25 18:23:32 myserver kernel: ffff88041bfdfdb8 0000000000000082 ffff88042be0bec0 ffff88041bfdffd8
Aug 25 18:23:32 myserver kernel: ffff88041bfdffd8 ffff88041bfdffd8 ffff88042be0bec0 ffff88042be0bec0
Aug 25 18:23:32 myserver kernel: ffff88042bd2b8b8 ffff88042bd2b8c0 ffffffff00000000 ffff88042bd2b8c8
Aug 25 18:23:32 myserver kernel: Call Trace:
Aug 25 18:23:32 myserver kernel: [<ffffffff8168c7f9>] schedule+0x29/0x70
Aug 25 18:23:32 myserver kernel: [<ffffffff8168dfa5>] rwsem_down_write_failed+0x115/0x220
Aug 25 18:23:32 myserver kernel: [<ffffffff81327647>] call_rwsem_down_write_failed+0x17/0x30
Aug 25 18:23:32 myserver kernel: [<ffffffff812a84c0>] ? cap_mmap_addr+0x60/0x60
Aug 25 18:23:32 myserver kernel: [<ffffffff8168b9bd>] down_write+0x2d/0x30
Aug 25 18:23:32 myserver kernel: [<ffffffff811a07fc>] vm_mmap_pgoff+0x8c/0xe0
Aug 25 18:23:32 myserver kernel: [<ffffffff811b62d6>] SyS_mmap_pgoff+0x116/0x270
Aug 25 18:23:32 myserver kernel: [<ffffffff8102fb82>] SyS_mmap+0x22/0x30
Aug 25 18:23:32 myserver kernel: [<ffffffff81697809>] system_call_fastpath+0x16/0x1b

Aqui estão algumas informações sobre o meu servidor, se forem úteis:

Centos 7.3 3.10.0-514.26.2.el7.x86_64 # 1 SMP Ter Jul 4 15:04:05 UTC 2017 x86_64 x86_64 x86_64 GNU / Linux - Quad core com 16 GB de RAM - Unidades RAID suaves de 2 TB

    

Answer 1

Confira este artigo sobre o impacto no desempenho do uso de imjournal:
link

Especialmente esta parte:

Warning: Some versions of systemd journal have problems with database corruption, which leads to the journal to return the same data endlessly in a tight loop. This results in massive message duplication inside rsyslog probably resulting in a denial-of-service when the system ressouces get exhausted. This can be somewhat mitigated by using proper rate-limiters, but even then there are spikes of old data which are endlessly repeated. By default, ratelimiting is activated and permits to process 20,000 messages within 10 minutes, what should be well enough for most use cases. If insufficient, use the parameters described below to adjust the permitted volume. It is strongly recommended to use this plugin only if there is hard need to do so.

Em suma, acho que você deve considerar usar o imuxsock.