[DRBD-user] Enabling a stacked resource - HELP!

SlingPirate sling.temp at googlemail.com
Thu Sep 8 17:47:12 CEST 2011

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


If I want to run VMs on both node 1 & node 2 they would need to be setup as
dual primaries, is this not the case?

What I was trying to achive was two on-site servers that could fail over to
one another OR in case both die node 3 runs all the servers (rather slowly
of course!)

Sorry for my ignorance but I dont quite understand your diagram for the
alternative.


Lars Ellenberg wrote:
> 
> On Thu, Sep 08, 2011 at 04:52:28PM +0200, Florian Haas wrote:
>> On 09/08/11 15:06, SlingPirate wrote:
>> > 
>> > I have 3 servers two are to run in dual primary mode (node 1 and 2)
>> while a
>> > third (node 3) is to be an off-site backup.  I have got the first 2
>> nodes
>> > sync'd and in primary mode quite happily but I can't get the backup
>> node to
>> > sync. Every time I try to bring the stacked resource into primary on
>> node 1
>> > (even with node 2 in secondary) it complains that it "needs access to
>> > UpToDate data".
>> > 
>> > Am I just doing things wrong by trying to do a dual primary with a
>> stacked
>> > backup?
>> 
>> Yes. This can't work.
> 
> Well. You can do this:
> 
> [ upper Primary node 1 ] ---- [ upper Primary node 2 ]
> [ lower DRBD node 1 ]
>     `------.[ DR node ]
> 
> And live with the fact that, if you lose node 1, you lose your DR link.
> If you want to resume the DR replication from node 2,
> you need a full sync to the DR site.
> 
> [If you do a planed migration of the DR link to node 2 (no crash
> involved), you can do so without a full sync to the DR site, but you'd
> need to know more DRBD internals than are good for me :-)]
> 
> If you lose node 1 catastrophically, and you now need to resume the DR
> replication from node 2, you will still need a full sync to the DR site.
> 
> 
> 
> This can not work:
> [ upper Primary node 1 ]------------------------------------------ [DR ]
> [ lower Primary node 1 ]-----[ lower Primary node 2 ]
> 
> Because there is no way for writes from node2 to reach DR in this
> picture, [DR] would receive only half of the writes, thus have a
> corrupted dataset by design.
> 
> 
> Maybe you don't need dual primary anyways?
> Very few installations actually do.
> 
> -- 
> : Lars Ellenberg
> : LINBIT | Your Way to High Availability
> : DRBD/HA support and consulting http://www.linbit.com
> 
> DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.
> __
> please don't Cc me, but send to list   --   I'm subscribed
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
> 
> 

-- 
View this message in context: http://old.nabble.com/Enabling-a-stacked-resource---HELP%21-tp32423513p32424927.html
Sent from the DRBD - User mailing list archive at Nabble.com.




More information about the drbd-user mailing list