You are viewing a plain text version of this content. The canonical link for it is here.

Posted to hdfs-dev@hadoop.apache.org by Brahma Reddy Battula <br...@huawei.com> on 2017/08/01 13:20:24 UTC

When permission is disabled, why setOwner() && setPermission() still check the permission?

Hi All

why  the "dfs.permissions.enabled" flag was not considered for this setOwner() and setPermission() check..?

and why extra super user required for setOwner()? is the check required even in case of permissions disabled?

Any idea on this..?



Thanks
Brahma Reddy Battula

RE: When permission is disabled, why setOwner() && setPermission() still check the permission?

Posted by Brahma Reddy Battula <br...@huawei.com>.

Thanks Vinay and Ravi for reply.

Yes, since branch-1 behavior is same. Would like to know whether it’s missed/purposefully kept?

If it is not intended, Shall I raise Jira and remove this permission checks..?

--Brahma Reddy Battula

From: Ravi Prakash [mailto:ravihadoop@gmail.com]
Sent: 09 August 2017 07:38
To: Vinayakumar B
Cc: Brahma Reddy Battula; hdfs-dev@hadoop.apache.org
Subject: Re: When permission is disabled, why setOwner() && setPermission() still check the permission?

I agree! I don't know of any one who doesn't use permissions. This was probably a feature flag that never got removed.

On Tue, Aug 8, 2017 at 3:29 AM, Vinayakumar B <vi...@apache.org>> wrote:
Hi Brahma,

Thanks for bringing this up.

According to history of git, there is no historical reasons found for me.
May be someone with branch-1 experience, can answer this ..?

-Vinay

On Tue, Aug 1, 2017 at 6:50 PM, Brahma Reddy Battula <
brahmareddy.battula@huawei.com<ma...@huawei.com>> wrote:

> Hi All
>
> why  the "dfs.permissions.enabled" flag was not considered for this
> setOwner() and setPermission() check..?
>
> and why extra super user required for setOwner()? is the check required
> even in case of permissions disabled?
>
> Any idea on this..?
>
>
>
> Thanks
> Brahma Reddy Battula
>
>

Re: When permission is disabled, why setOwner() && setPermission() still check the permission?

Posted by Ravi Prakash <ra...@gmail.com>.

I agree! I don't know of any one who doesn't use permissions. This was
probably a feature flag that never got removed.

On Tue, Aug 8, 2017 at 3:29 AM, Vinayakumar B <vi...@apache.org>
wrote:

> Hi Brahma,
>
> Thanks for bringing this up.
>
> According to history of git, there is no historical reasons found for me.
> May be someone with branch-1 experience, can answer this ..?
>
> -Vinay
>
> On Tue, Aug 1, 2017 at 6:50 PM, Brahma Reddy Battula <
> brahmareddy.battula@huawei.com> wrote:
>
> > Hi All
> >
> > why  the "dfs.permissions.enabled" flag was not considered for this
> > setOwner() and setPermission() check..?
> >
> > and why extra super user required for setOwner()? is the check required
> > even in case of permissions disabled?
> >
> > Any idea on this..?
> >
> >
> >
> > Thanks
> > Brahma Reddy Battula
> >
> >
>

Re: When permission is disabled, why setOwner() && setPermission() still check the permission?

Posted by Vinayakumar B <vi...@apache.org>.

Hi Brahma,

Thanks for bringing this up.

According to history of git, there is no historical reasons found for me.
May be someone with branch-1 experience, can answer this ..?

-Vinay

On Tue, Aug 1, 2017 at 6:50 PM, Brahma Reddy Battula <
brahmareddy.battula@huawei.com> wrote:

> Hi All
>
> why  the "dfs.permissions.enabled" flag was not considered for this
> setOwner() and setPermission() check..?
>
> and why extra super user required for setOwner()? is the check required
> even in case of permissions disabled?
>
> Any idea on this..?
>
>
>
> Thanks
> Brahma Reddy Battula
>
>