不确定是否存在 ZFS 问题/ATA Linux 控制器问题?

不确定是否存在 ZFS 问题/ATA Linux 控制器问题?

我已经使用 ZFS 几年了,但自从转移到新的 NAS 后,我遇到了一个奇怪的问题,我甚至不确定是否与 ZFS 有关。

我运行一个带有 4 个 4TB 磁盘的 RAID-Z,大约每周(有时每 2/3 天)我的dmesg输出开始显示:

[181263.131531]  sde: sde1 sde9
[181264.346713]  sdc: sdc1 sdc9
[181265.722670]  sdb: sdb1 sdb9
[181267.026973]  sdd: sdd1 sdd9
[181269.223494]  sde: sde1 sde9
[181270.562879]  sdb: sdb1 sdb9
[181271.955974]  sdd: sdd1 sdd9
[181273.965705]  sde: sde1 sde9
[181275.147701]  sdb: sdb1 sdb9
[181276.085743]  sdd: sdd1 sdd9
[181276.880991]  sde: sde1 sde9
[181278.314671]  sde: sde1 sde9
[181309.204054]  sde: sde1 sde9

一遍又一遍,我的池开始重新同步,有时会发现校验和错误,有时根本没有发现任何错误。我使用了 B450I Aorus Pro ITX 主板上的板载 SATA 端口,但最近得到了一个带有 6 个端口的 ASMedia 109X 廉价 PCIe 控制器,得到了相同的结果。尝试使用内核 5.4.x 和 5.3.x,没有区别。

NAS 运行基于 Debian 10 的 OpenMediaVault 5.x。CPU 是 AMG 200GE。

欢迎任何帮助。谢谢!

更新:这些是内核记录器中紧接着之前的几行:

[  541.834989] docker0: port 16(veth89a0c65) entered blocking state
[  541.834992] docker0: port 16(veth89a0c65) entered forwarding state
[19666.977275] perf: interrupt took too long (2505 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
[65299.449905] docker0: port 17(veth342b51a) entered blocking state
[65299.449908] docker0: port 17(veth342b51a) entered disabled state
[65299.450102] device veth342b51a entered promiscuous mode
[65299.478322] docker0: port 18(veth2afecda) entered blocking state
[65299.478326] docker0: port 18(veth2afecda) entered disabled state
[65299.478433] device veth2afecda entered promiscuous mode
[65299.478566] docker0: port 18(veth2afecda) entered blocking state
[65299.478568] docker0: port 18(veth2afecda) entered forwarding state
[65299.840540] eth0: renamed from vethc2c1575
[65299.855726] docker0: port 18(veth2afecda) entered disabled state
[65299.855786] docker0: port 17(veth342b51a) entered blocking state
[65299.855787] docker0: port 17(veth342b51a) entered forwarding state
[65300.287422] eth0: renamed from vethcda7502
[65300.303228] docker0: port 18(veth2afecda) entered blocking state
[65300.303230] docker0: port 18(veth2afecda) entered forwarding state
[181263.131531]  sde: sde1 sde9
[181264.346713]  sdc: sdc1 sdc9
[181265.722670]  sdb: sdb1 sdb9
[181267.026973]  sdd: sdd1 sdd9
[181269.223494]  sde: sde1 sde9
[181270.562879]  sdb: sdb1 sdb9
[181271.955974]  sdd: sdd1 sdd9
[181273.965705]  sde: sde1 sde9
[181275.147701]  sdb: sdb1 sdb9
[181276.085743]  sdd: sdd1 sdd9
[181276.880991]  sde: sde1 sde9
[181278.314671]  sde: sde1 sde9
[181309.204054]  sde: sde1 sde9

以及其他相关内容sdb

[    3.900995] sd 4:0:0:0: [sdb] 7814037168 512-byte logical blocks: (4.00 TB/3.64 TiB)
[    3.900997] sd 4:0:0:0: [sdb] 4096-byte physical blocks
[    3.901022] sd 4:0:0:0: [sdb] Write Protect is off
[    3.901024] sd 4:0:0:0: [sdb] Mode Sense: 00 3a 00 00
[    3.901059] sd 4:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[    3.974764]  sdb: sdb1 sdb9
[    3.975748] sd 4:0:0:0: [sdb] Attached SCSI disk
[181265.722670]  sdb: sdb1 sdb9
[181270.562879]  sdb: sdb1 sdb9
[181275.147701]  sdb: sdb1 sdb9

所以看起来驱动器枚举线似乎是凭空而来的……?

相关内容