Re: MD Array Unexpected Kernel Hang

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hey all,

Thanks for the responses!

> A quick google also says these are old drives, which may or not be a 
> concern. That also possibly explains the lack of sct/erc.

These are old SAS drives, correct.  They're in an external JBOD chassis with an HBA expander to the host machine.  I acquired them used with the hopes of extending their life in a raid array.

> Given that you say three drives all failed in the first slot? My money 
> would actually be on nothing to do with raid, but a dodgy cable or 
> motherboard connector. I don't think they're rated at being swapped over 
> that many times ...

First position in the virtual array, not physical.  The actual drives are plugged into different drive bays and I'm swapping them into the VM by updating the disk's libvirt XML configuration.

> I don't know how much help this website will be for you, but take a look...
>
> https://raid.wiki.kernel.org/index.php/Linux_Raid

Thanks!  I've spent a lot of time on that wiki, it's very helpful!  It's also why I've come to the mailing list for help, since I've been using md arrays since 2015 and haven't encountered an issue like this before.

> I would question why he is passing the disks in via virtio, rather than doing the raid outside the VM and passing in the raid6 block device.

I do it for the convenience of managing VMs.  It's for a homelab, so I won't always have remote hands to replace a disk and I like to have the spares free to use for other projects instead of tying them up as a hot-spare full-time.

> I don't believe that using MD-raid with devices virtualized and passed through via a virtio device is going to be valid/supportable.

I realize it's a bit wonky, but I'm surprised that this would be the case.  I wouldn't expect it to be that uncommon these days.  I guess I can always go back to zfs if you all feel like this is too crazy. :-P

If anybody has any other tips for debugging, it would be much appreciated.  I feel that ebpf might be helpful here, but haven't yet been able to quite figure out the right way to inspect the situation with BCC-tools.

Thanks everyone,
- Alan



[Index of Archives]     [Linux RAID Wiki]     [ATA RAID]     [Linux SCSI Target Infrastructure]     [Linux Block]     [Linux IDE]     [Linux SCSI]     [Linux Hams]     [Device Mapper]     [Device Mapper Cryptographics]     [Kernel]     [Linux Admin]     [Linux Net]     [GFS]     [RPM]     [git]     [Yosemite Forum]


  Powered by Linux