All of lore.kernel.org
 help / color / mirror / Atom feed
* seagate hm-smr issue
@ 2021-09-26 13:57 Jingyun He
  2021-09-26 23:13 ` Damien Le Moal
  0 siblings, 1 reply; 5+ messages in thread
From: Jingyun He @ 2021-09-26 13:57 UTC (permalink / raw)
  To: linux-btrfs

Hello,
Btrfs works very well on WD/HGST disks, we got some Seagate HM-SMR
disks recently,  model number is ST14000NM0428,
mkfs.btrfs works fine, and I can mount it, and push data into disk.
once we used up the capacity, and umount it. then we are unable to
re-mount it again.
The mount process will never end, the process just hangs there.

Anybody can help me with this?

Thanks.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: seagate hm-smr issue
  2021-09-26 13:57 seagate hm-smr issue Jingyun He
@ 2021-09-26 23:13 ` Damien Le Moal
  2021-09-26 23:42   ` Jingyun He
  0 siblings, 1 reply; 5+ messages in thread
From: Damien Le Moal @ 2021-09-26 23:13 UTC (permalink / raw)
  To: Jingyun He, linux-btrfs, Naohiro Aota, Johannes Thumshirn

On 2021/09/26 22:57, Jingyun He wrote:
> Hello,
> Btrfs works very well on WD/HGST disks, we got some Seagate HM-SMR
> disks recently,  model number is ST14000NM0428,
> mkfs.btrfs works fine, and I can mount it, and push data into disk.
> once we used up the capacity, and umount it. then we are unable to
> re-mount it again.
> The mount process will never end, the process just hangs there.
> 
> Anybody can help me with this?

+Naohiro and Johannes

This is not a hang. Mount will just take a looooong time due to an inefficiency
in how block groups are checked: a single zone report zone command is issued per
block group, so with a disk almost full, that takes a long time (75000+zones
checked one by one). Naohiro is working on a fix for this.

> 
> Thanks.
> 


-- 
Damien Le Moal
Western Digital Research

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: seagate hm-smr issue
  2021-09-26 23:13 ` Damien Le Moal
@ 2021-09-26 23:42   ` Jingyun He
  2021-09-27  0:07     ` Damien Le Moal
  0 siblings, 1 reply; 5+ messages in thread
From: Jingyun He @ 2021-09-26 23:42 UTC (permalink / raw)
  To: Damien Le Moal; +Cc: linux-btrfs, Naohiro Aota, Johannes Thumshirn

Hi,
You are correct, for a WD/HGST HM-SMR disk, it takes about 15 mins to
mount a full disk.

But It takes about 2-3 hours for mounting a Seagate HM-SMR disk.

Have no idea why Seagate takes so loooooong time.

On Mon, Sep 27, 2021 at 7:13 AM Damien Le Moal
<damien.lemoal@opensource.wdc.com> wrote:
>
> On 2021/09/26 22:57, Jingyun He wrote:
> > Hello,
> > Btrfs works very well on WD/HGST disks, we got some Seagate HM-SMR
> > disks recently,  model number is ST14000NM0428,
> > mkfs.btrfs works fine, and I can mount it, and push data into disk.
> > once we used up the capacity, and umount it. then we are unable to
> > re-mount it again.
> > The mount process will never end, the process just hangs there.
> >
> > Anybody can help me with this?
>
> +Naohiro and Johannes
>
> This is not a hang. Mount will just take a looooong time due to an inefficiency
> in how block groups are checked: a single zone report zone command is issued per
> block group, so with a disk almost full, that takes a long time (75000+zones
> checked one by one). Naohiro is working on a fix for this.
>
> >
> > Thanks.
> >
>
>
> --
> Damien Le Moal
> Western Digital Research

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: seagate hm-smr issue
  2021-09-26 23:42   ` Jingyun He
@ 2021-09-27  0:07     ` Damien Le Moal
  2021-12-27 17:46       ` Jingyun He
  0 siblings, 1 reply; 5+ messages in thread
From: Damien Le Moal @ 2021-09-27  0:07 UTC (permalink / raw)
  To: Jingyun He; +Cc: linux-btrfs, Naohiro Aota, Johannes Thumshirn

On 2021/09/27 8:42, Jingyun He wrote:
> Hi,
> You are correct, for a WD/HGST HM-SMR disk, it takes about 15 mins to
> mount a full disk.
> 
> But It takes about 2-3 hours for mounting a Seagate HM-SMR disk.
> 
> Have no idea why Seagate takes so loooooong time.

This is most likely due to a difference in the execution time of report zones
commands.

In any case, this will be resolved soon as we can cache a full disk zone report
during the mount operation to avoid this slow per-zone report operation.

> 
> On Mon, Sep 27, 2021 at 7:13 AM Damien Le Moal
> <damien.lemoal@opensource.wdc.com> wrote:
>>
>> On 2021/09/26 22:57, Jingyun He wrote:
>>> Hello,
>>> Btrfs works very well on WD/HGST disks, we got some Seagate HM-SMR
>>> disks recently,  model number is ST14000NM0428,
>>> mkfs.btrfs works fine, and I can mount it, and push data into disk.
>>> once we used up the capacity, and umount it. then we are unable to
>>> re-mount it again.
>>> The mount process will never end, the process just hangs there.
>>>
>>> Anybody can help me with this?
>>
>> +Naohiro and Johannes
>>
>> This is not a hang. Mount will just take a looooong time due to an inefficiency
>> in how block groups are checked: a single zone report zone command is issued per
>> block group, so with a disk almost full, that takes a long time (75000+zones
>> checked one by one). Naohiro is working on a fix for this.
>>
>>>
>>> Thanks.
>>>
>>
>>
>> --
>> Damien Le Moal
>> Western Digital Research


-- 
Damien Le Moal
Western Digital Research

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: seagate hm-smr issue
  2021-09-27  0:07     ` Damien Le Moal
@ 2021-12-27 17:46       ` Jingyun He
  0 siblings, 0 replies; 5+ messages in thread
From: Jingyun He @ 2021-12-27 17:46 UTC (permalink / raw)
  To: Damien Le Moal; +Cc: linux-btrfs, Naohiro Aota, Johannes Thumshirn

Hello,
Do you have any update on this? or any patch that I can try?

Thanks.

On Mon, Sep 27, 2021 at 8:07 AM Damien Le Moal
<damien.lemoal@opensource.wdc.com> wrote:
>
> On 2021/09/27 8:42, Jingyun He wrote:
> > Hi,
> > You are correct, for a WD/HGST HM-SMR disk, it takes about 15 mins to
> > mount a full disk.
> >
> > But It takes about 2-3 hours for mounting a Seagate HM-SMR disk.
> >
> > Have no idea why Seagate takes so loooooong time.
>
> This is most likely due to a difference in the execution time of report zones
> commands.
>
> In any case, this will be resolved soon as we can cache a full disk zone report
> during the mount operation to avoid this slow per-zone report operation.
>
> >
> > On Mon, Sep 27, 2021 at 7:13 AM Damien Le Moal
> > <damien.lemoal@opensource.wdc.com> wrote:
> >>
> >> On 2021/09/26 22:57, Jingyun He wrote:
> >>> Hello,
> >>> Btrfs works very well on WD/HGST disks, we got some Seagate HM-SMR
> >>> disks recently,  model number is ST14000NM0428,
> >>> mkfs.btrfs works fine, and I can mount it, and push data into disk.
> >>> once we used up the capacity, and umount it. then we are unable to
> >>> re-mount it again.
> >>> The mount process will never end, the process just hangs there.
> >>>
> >>> Anybody can help me with this?
> >>
> >> +Naohiro and Johannes
> >>
> >> This is not a hang. Mount will just take a looooong time due to an inefficiency
> >> in how block groups are checked: a single zone report zone command is issued per
> >> block group, so with a disk almost full, that takes a long time (75000+zones
> >> checked one by one). Naohiro is working on a fix for this.
> >>
> >>>
> >>> Thanks.
> >>>
> >>
> >>
> >> --
> >> Damien Le Moal
> >> Western Digital Research
>
>
> --
> Damien Le Moal
> Western Digital Research

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2021-12-27 17:47 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-26 13:57 seagate hm-smr issue Jingyun He
2021-09-26 23:13 ` Damien Le Moal
2021-09-26 23:42   ` Jingyun He
2021-09-27  0:07     ` Damien Le Moal
2021-12-27 17:46       ` Jingyun He

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.