Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 28 Dec 2023 09:39:31 +1300
From:      Jonathan Chen <jonc@chen.org.nz>
To:        stable@freebsd.org
Subject:   Re: Unusual ZFS behaviour
Message-ID:  <735d5bc7-fc83-4470-abde-2dca43bf7496@chen.org.nz>
In-Reply-To: <3cad03f3-f3f6-4d6f-976b-55f5af8909db@chen.org.nz>
References:  <f8764549-773a-4695-b1fc-76e70e49de1b@chen.org.nz> <3cad03f3-f3f6-4d6f-976b-55f5af8909db@chen.org.nz>

next in thread | previous in thread | raw e-mail | index | archive | help
On 23/11/23 19:06, Jonathan Chen wrote:
> On 22/11/23 19:49, Jonathan Chen wrote:
>> Hi,
>>
>> I'm running a somewhat recent version of STABLE-13/amd64: 
>> stable/13-n256681-0b7939d725ba: Fri Nov 10 08:48:36 NZDT 2023, and I'm 
>> seeing some unusual behaviour with ZFS.
>>
>> To reproduce:
>>   1. one big empty disk, GPT scheme, 1 freebsd-zfs partition.
>>   2. create a zpool, eg: tank
>>   3. create 2 sub-filesystems, eg: tank/one, tank/two
>>   4. fill each sub-filesystem with large files until the pool is ~80% 
>> full. In my case I had 200 10Gb files in each.
>>   5. in one session run 'md5 tank/one/*'
>>   6. in another session run 'md5 tank/two/*'
>>
>> For most of my runs, one of the sessions against a sub-filesystem will 
>> be starved of I/O, while the other one is performant.
> 
> I've run a few more tests, and the issue appears to be isolated to my 
> Alder Lake based system only. So it's more likely to be an issue with 
> the 'Alder Lake-S PCH SATA Controller [AHCI Mode]' or maybe the 
> scheduler using the P & E cores.

I've updated to FreeBSD 14/STABLE, and I'm glad to report that this bug 
has gone away. I personally suspect that:
   https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=274698
is the cause.

Cheers.
-- 
Jonathan Chen <jonc@chen.org.nz>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?735d5bc7-fc83-4470-abde-2dca43bf7496>