linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Tkaczyk, Mariusz" <mariusz.tkaczyk@linux.intel.com>
To: Xiao Ni <xni@redhat.com>, linux-raid <linux-raid@vger.kernel.org>,
	artur.paszkiewicz@intel.com,
	Jes Sorensen <jes.sorensen@gmail.com>, NeilBrown <neilb@suse.com>,
	Nigel Croxon <ncroxon@redhat.com>
Subject: Re: One failed raid device can't umount automatically
Date: Thu, 4 Feb 2021 09:25:24 +0100	[thread overview]
Message-ID: <fc497639-925d-c217-e241-52f293eea382@linux.intel.com> (raw)
In-Reply-To: <b4cc93d4-4923-4959-3258-f03eca58f18e@redhat.com>

On 01.02.2021 15:35, Xiao Ni wrote:
> Hi all
Hi Xiao,
> 
> Any good suggestion for this problem?
> 
Replacing udisk by umount won't work from udev context. Umount gets
perrmision denied. You should ask systemd developers to find out
udev friendly way for stopping from udev context. If it is not possible,
in my opinion we should drop this functionality.

> Regards
> Xiao
> 
> On 01/12/2021 04:42 PM, Xiao Ni wrote:
>> Hi all
>>
>> We support to umount one failed raid device automatically. But it can't work now.
>> For example, one 3 disks raid5 device /dev/md0. I unplug two disks one by one.
>> The udev rule udev-md-raid-assembly.rules is triggered when unplug disk.
>>
>> In this udev rule, it calls `mdadm -If $disk` when unplug one disk. Function 
>> IncrementalRemove
>> is called. When the raid doesn't have enough disks to be active, it tries to 
>> stop the array.
>> Before stopping the array, it tries to umount the raid device first.
>>
>> Now it uses udisks to umount raid device. I printed logs during test. It gives 
>> error message
>> "Permission denied". Then I tried with umount directly, it failed with the 
>> same error message.
>>
>> diff --git a/Incremental.c b/Incremental.c
>> index e849bdd..96ba234 100644
>> --- a/Incremental.c
>> +++ b/Incremental.c
>> @@ -1620,6 +1620,7 @@ static void run_udisks(char *arg1, char *arg2)
>>                 manage_fork_fds(1);
>>                 execl("/usr/bin/udisks", "udisks", arg1, arg2, NULL);
>>                 execl("/bin/udisks", "udisks", arg1, arg2, NULL);
>> +               execl("/usr/bin/umount", "umount", arg2, NULL);
>>                 exit(1);
>>         }
>>         while (pid > 0 && wait(&status) != pid)
>>
>> Does anyone know how to fix this problem?
>>
>> Regards
>> Xiao
>>
>
Thanks,
Mariusz

      reply	other threads:[~2021-02-04  8:27 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-01-12  8:42 One failed raid device can't umount automatically Xiao Ni
2021-02-01 14:35 ` Xiao Ni
2021-02-04  8:25   ` Tkaczyk, Mariusz [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=fc497639-925d-c217-e241-52f293eea382@linux.intel.com \
    --to=mariusz.tkaczyk@linux.intel.com \
    --cc=artur.paszkiewicz@intel.com \
    --cc=jes.sorensen@gmail.com \
    --cc=linux-raid@vger.kernel.org \
    --cc=ncroxon@redhat.com \
    --cc=neilb@suse.com \
    --cc=xni@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).