내 홈 서버(Debian 11, 커널 5.17.1)는 이전 Synology DiskStation DS409+(커널 2.6.32.12)에서 NFS v3을 통해 네트워크 공유를 마운트하고 매주 백업을 수행합니다.
백업 도구로 사용하고 있는데 rsnapshot여러 원격 서버의 파일을 백업하기 때문에 백업 과정이 꽤 오랜 시간(수 시간)이 걸립니다.
약 한 시간 후에 NFS 문제가 나타났고 dmesg, 첫 번째 시간 초과 후 12분 후에 커널에 오류가 발생했습니다.
(backup starts at 06:10)
[Mo Jan 2 07:02:32 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:32 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:32 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:32 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:02:37 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:05:33 2023] rpc_check_timeout: 4654 callbacks suppressed
[Mo Jan 2 07:05:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:05:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:08:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:08:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:11:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:11:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:11:33 2023] nfs: server disa not responding, timed out
[Mo Jan 2 07:14:29 2023] INFO: task rsync:63109 blocked for more than 120 seconds.
[Mo Jan 2 07:14:29 2023] Tainted: G E 5.17.1 #1
[Mo Jan 2 07:14:29 2023] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[Mo Jan 2 07:14:29 2023] task:rsync state:D stack: 0 pid:63109 ppid: 63107 flags:0x00004000
[Mo Jan 2 07:14:29 2023] Call Trace:
[Mo Jan 2 07:14:29 2023] <TASK>
[Mo Jan 2 07:14:29 2023] __schedule+0x2e5/0x9b0
[Mo Jan 2 07:14:29 2023] schedule+0x4e/0xc0
[Mo Jan 2 07:14:29 2023] io_schedule+0x3b/0x60
[Mo Jan 2 07:14:29 2023] folio_wait_bit+0x101/0x230
[Mo Jan 2 07:14:29 2023] ? find_get_pages_range_tag+0x1a0/0x1f0
[Mo Jan 2 07:14:29 2023] ? filemap_alloc_folio+0xb0/0xb0
[Mo Jan 2 07:14:29 2023] folio_wait_writeback+0x28/0x80
[Mo Jan 2 07:14:29 2023] __filemap_fdatawait_range+0x83/0xf0
[Mo Jan 2 07:14:29 2023] filemap_write_and_wait_range+0x81/0xd0
[Mo Jan 2 07:14:29 2023] nfs_wb_all+0x22/0x120 [nfs]
[Mo Jan 2 07:14:29 2023] nfs_file_write+0x250/0x2a0 [nfs]
[Mo Jan 2 07:14:29 2023] new_sync_write+0x122/0x1b0
[Mo Jan 2 07:14:29 2023] vfs_write+0x1bb/0x2a0
[Mo Jan 2 07:14:29 2023] ksys_write+0x5f/0xe0
[Mo Jan 2 07:14:29 2023] do_syscall_64+0x3b/0xc0
[Mo Jan 2 07:14:29 2023] entry_SYSCALL_64_after_hwframe+0x44/0xae
[Mo Jan 2 07:14:29 2023] RIP: 0033:0x7fa59bae88f3
[Mo Jan 2 07:14:29 2023] RSP: 002b:00007ffcf410a958 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[Mo Jan 2 07:14:29 2023] RAX: ffffffffffffffda RBX: 000055ce885fd920 RCX: 00007fa59bae88f3
[Mo Jan 2 07:14:29 2023] RDX: 0000000000040000 RSI: 000055ce885fd920 RDI: 0000000000000004
[Mo Jan 2 07:14:29 2023] RBP: 0000000000000004 R08: 0000000000000003 R09: 0000000000000000
[Mo Jan 2 07:14:29 2023] R10: 0000000000000000 R11: 0000000000000246 R12: 000055ce87a2d329
[Mo Jan 2 07:14:29 2023] R13: 0000000000000004 R14: 0000000000000005 R15: 00007ffcf410aa28
[Mo Jan 2 07:14:29 2023] </TASK>
[Mo Jan 2 07:14:33 2023] nfs: server disa not responding, timed out
(timeouts continue forever)
의 설치 옵션은 /etc/fstab다음과 같습니다.user,nolock,nfsvers=4,soft
"mount"에 대해 나열된 실제 유효한 옵션은 다음과 같습니다.
rw
nosuid
nodev
noexec
relatime
vers=3
rsize=131072
wsize=131072
namlen=255
soft
nolock
proto=tcp
timeo=600 - time in deciseconds (tenths of a second)
retrans=2
sec=sys
mountaddr=192.168.3.96
mountvers=3
mountport=892
mountproto=udp
local_lock=all
addr=192.168.3.96
user
NAS에서 오류 로그 메시지를 찾을 수 없습니다.
NAS가 오래되었고 2016년 이후로 업데이트되지 않았다는 것을 알고 있습니다(그리고 "실제" Linux를 설치하는 방법도 모릅니다). 그러나 하드웨어는 여전히 잘 작동하므로 계속 사용하고 싶습니다.
시간이 초과되는 이유는 무엇입니까? 시간 초과를 방지하려면 어떻게 해야 합니까?