On Mon, Aug 21, 2017 at 3:03 PM, Bryan Banister
<[email protected]> wrote:
> Thanks for the response John.
>
> Maybe I'm not understanding this correctly, but I thought clients could be 
> restricted to specific file systems by limiting access to the underlying ceph 
> pools used in each file system?
>
> client.cephfs.test1
>         key: AQDuQpdZp90MHhAAkYE6P5XYzsoswgEkZy6RLw==
>         caps: [mds] allow
>         caps: [mon] allow r
>         caps: [osd] allow rw pool cephfs01_data
>
> client.cephfs.test2
>         key: AQDuQpdZp90MHhAAkYE6P5XYzsoswgEkZy6RLw==
>         caps: [mds] allow
>         caps: [mon] allow r
>         caps: [osd] allow rw pool cephfs02_data
>
> Would these two clients keys which only have access to specific data pools 
> restrict their access?
>
> Or I guess with mds allow on both, then they could mount the file system but 
> only reading/writing the data in the file systems would be restricted?

Correct, although critically even if a client doesn't have pool caps,
it can still send *deletes* to the MDS and thereby attack the data.

In the long run the caps should be improved so that we are always
giving a client access to a specific filesystem, so that these sorts
of fine lines just go away.

John

>
> Thanks!
> -Bryan
>
> -----Original Message-----
> From: John Spray [mailto:[email protected]]
> Sent: Monday, August 21, 2017 8:48 AM
> To: Bryan Banister <[email protected]>
> Cc: [email protected]
> Subject: Re: [ceph-users] Any experience with multiple cephfs instances in 
> one ceph cluster? How experimental is this?
>
> Note: External Email
> -------------------------------------------------
>
> On Mon, Aug 21, 2017 at 2:35 PM, Bryan Banister
> <[email protected]> wrote:
>> Hi all,
>>
>>
>>
>> I’m very new to ceph and cephfs, so I’m just starting to play around with
>> the Luminous release.  There are some very concerning warnings about
>> deploying multiple cephfs instances in the same cluster:
>>
>> “There are no known bugs, but any failures which do result from having
>> multiple active filesystems in your cluster will require manual intervention
>> and, so far, will not have been experienced by anybody else – knowledgeable
>> help will be extremely limited. You also probably do not have the security
>> or isolation guarantees you want or think you have upon doing so.”
>
> The sort of literal translation of this is:
>  - the automated tests for systems with multiple filesystems are not
> very comprehensive
>  - a client that can access one filesystem can access all of them
>
> If you're adventurous enough to be running upstream Ceph packages, and
> you have at least some level of test/staging environment to try it in,
> then I'd not be too scared about trying it out.
>
>> And Redhat says:
>>
>> “Creating multiple Ceph File Systems in one cluster is not fully supported
>> yet and can cause the MDS or client nodes to terminate unexpectedly.”
>
> I don't know who wrote that text, but I do not believe that there are
> any known issues involving MDS or client nodes terminating
> unexpectedly.
>
> John
>
>>
>>
>>
>> Is anybody deploying multiple cephfs instances and have there been any
>> issues like the warnings indicate?
>>
>>
>>
>> Thanks!
>>
>> -Bryan
>>
>>
>>
>>
>>
>>
>> ________________________________
>>
>> Note: This email is for the confidential use of the named addressee(s) only
>> and may contain proprietary, confidential or privileged information. If you
>> are not the intended recipient, you are hereby notified that any review,
>> dissemination or copying of this email is strictly prohibited, and to please
>> notify the sender immediately and destroy this email and any attachments.
>> Email transmission cannot be guaranteed to be secure or error-free. The
>> Company, therefore, does not make any guarantees as to the completeness or
>> accuracy of this email or any attachments. This email is for informational
>> purposes only and does not constitute a recommendation, offer, request or
>> solicitation of any kind to buy, sell, subscribe, redeem or perform any type
>> of transaction of a financial product.
>>
>> _______________________________________________
>> ceph-users mailing list
>> [email protected]
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>
>
> ________________________________
>
> Note: This email is for the confidential use of the named addressee(s) only 
> and may contain proprietary, confidential or privileged information. If you 
> are not the intended recipient, you are hereby notified that any review, 
> dissemination or copying of this email is strictly prohibited, and to please 
> notify the sender immediately and destroy this email and any attachments. 
> Email transmission cannot be guaranteed to be secure or error-free. The 
> Company, therefore, does not make any guarantees as to the completeness or 
> accuracy of this email or any attachments. This email is for informational 
> purposes only and does not constitute a recommendation, offer, request or 
> solicitation of any kind to buy, sell, subscribe, redeem or perform any type 
> of transaction of a financial product.
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to