Lustre discuss - Jun 2006 - how to backup the MDS data to other server(passive) disk

Hi All,
       In our testing environment we have 

   2 - OSS (each one serving 1 OST)           - 2 OST
   2 - MDS (active/passive, no shared storage)- 2 MDT

since we don''t have shared storage, how to backup the
MDT contents to passive server??, so that it can take
over active server in case of failure.

  pls suggest me if there is any better way to achieve
this, with out shared storage.

thanks in advance,
-bala-

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

Hi,

First of all, shared storage (I assume you mean to mount the MDT on
another server) wouldn''t necessarily be a safe solution.

A good way to do this is to run the MDT on LVM, make an LVM snapshot,
backup (fully or incrementally) the snapshot volume by mounting that as
ldiskfs.  Your backup tool would need to know about all extended ext3
attributes but there are no other special features in the MDT disk file
system.  The same holds for OST''s.

Thanks!

- Peter -
> -----Original Message-----
> From: lustre-discuss-bounces@clusterfs.com 
> [mailto:lustre-discuss-bounces@clusterfs.com] On Behalf Of Bala
> Sent: Monday, June 26, 2006 4:34 AM
> To: lustre-discuss@clusterfs.com
> Subject: [Lustre-discuss] how to backup the MDS data to 
> otherserver(passive) disk
> 
> Hi All,
>        In our testing environment we have 
> 
>    2 - OSS (each one serving 1 OST)           - 2 OST
>    2 - MDS (active/passive, no shared storage)- 2 MDT
> 
> since we don''t have shared storage, how to backup the MDT 
> contents to passive server??, so that it can take over active 
> server in case of failure.
> 
>   pls suggest me if there is any better way to achieve this, 
> with out shared storage.
> 
> thanks in advance,
> -bala-
> 
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection 
> around http://mail.yahoo.com 
> _______________________________________________
> Lustre-discuss mailing list
> Lustre-discuss@clusterfs.com
> https://mail.clusterfs.com/mailman/listinfo/lustre-discuss
> 
>

Of course, drbd is commonly recommended when you don''t have special 
hardware for shared storage, but you still want quick failover capability. 
See http://drbd.org

On Tue, 27 Jun 2006, Peter J. Braam wrote:
> Hi,
>
> First of all, shared storage (I assume you mean to mount the MDT on
> another server) wouldn''t necessarily be a safe solution.
>
> A good way to do this is to run the MDT on LVM, make an LVM snapshot,
> backup (fully or incrementally) the snapshot volume by mounting that as
> ldiskfs.  Your backup tool would need to know about all extended ext3
> attributes but there are no other special features in the MDT disk file
> system.  The same holds for OST''s.
>
> Thanks!
>
> - Peter -
>
>> -----Original Message-----
>> From: lustre-discuss-bounces@clusterfs.com
>> [mailto:lustre-discuss-bounces@clusterfs.com] On Behalf Of Bala
>> Sent: Monday, June 26, 2006 4:34 AM
>> To: lustre-discuss@clusterfs.com
>> Subject: [Lustre-discuss] how to backup the MDS data to
>> otherserver(passive) disk
>>
>> Hi All,
>>        In our testing environment we have
>>
>>    2 - OSS (each one serving 1 OST)           - 2 OST
>>    2 - MDS (active/passive, no shared storage)- 2 MDT
>>
>> since we don''t have shared storage, how to backup the MDT
>> contents to passive server??, so that it can take over active
>> server in case of failure.
>>
>>   pls suggest me if there is any better way to achieve this,
>> with out shared storage.
>>
>> thanks in advance,
>> -bala-
>>
>> __________________________________________________
>> Do You Yahoo!?
>> Tired of spam?  Yahoo! Mail has the best spam protection
>> around http://mail.yahoo.com
>> _______________________________________________
>> Lustre-discuss mailing list
>> Lustre-discuss@clusterfs.com
>> https://mail.clusterfs.com/mailman/listinfo/lustre-discuss
>>
>>
> _______________________________________________
> Lustre-discuss mailing list
> Lustre-discuss@clusterfs.com
> https://mail.clusterfs.com/mailman/listinfo/lustre-discuss
>