Re: [RFC PATCH 2/2] nvme-multipath: remove multipath module param

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

 




On 4/9/25 4:15 PM, Christoph Hellwig wrote:
> On Tue, Apr 08, 2025 at 08:05:20PM +0530, Nilay Shroff wrote:
>> Okay, we can add an option to avoid making this behavior "the default".
>> So do you recommend adding a module option for opting in this behavior 
>> change or something else?
> 
> I guess a module option as default makes sense.  I'd still love to figure
> out a way to have per-controller options of some kind as e.g. this
> option make very little sense for thunderbolt-attached external devices.
> 
> But unfortunately I'm a bit lost what a good interface for that would be.
> 
> 
I don't know how to make this option per-controller as you know the 
head node, typically, refers to namespace paths and each path then 
refers to different controller. So if we were to make this option
per controller then how could we handle it in case one controller has
this option set but then the another controller doesn't set this 
option. It could be confusing. 
 
How about module option "nvme_core.multipath_head_always"? The default is
set to false. So now it becomes two step process:
1. modprobe nvme_core multipath_head_always=Y && modprobe nvme
2. echo "<val>" > /sys/block/nvme0XnY/delayed_shutdown_sec     

Thanks,
--Nilay




[Index of Archives]     [Linux RAID]     [Linux SCSI]     [Linux ATA RAID]     [IDE]     [Linux Wireless]     [Linux Kernel]     [ATH6KL]     [Linux Bluetooth]     [Linux Netdev]     [Kernel Newbies]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Device Mapper]

  Powered by Linux