[DRBD-user] Ask some questions about the splitting of the dual master device

Su Hua suhua.tanke at gmail.com
Mon Dec 10 09:55:32 CET 2018

Hi Roland KammererBelow
This is our version information:

The arbitration service I am talking about means that you can decide
to have only one device to provide services before the brain split occurs.

We know that the drbd configuration file can set the data recovery policy
script that is executed after the brain split, but these strategies may
cause partial data loss.

like :
net {
        after-sb-0pri discard-younger-primary;

Therefore, in order to prevent data loss or IO tear, a stricter arbitration
strategy is needed to ensure that all IOs can only select one controller
to provide services after brain splitting.

Moreover, it is necessary to ensure that the IO data in the pending state
after the synchronization network port is broken is returned in an error
manner, and then the data is retransmitted to the controller providing
the service selected by the arbitration mechanism.

Roland Kammerer <roland.kammerer at linbit.com> 于2018年12月10日周一 下午3:14写道:

> Su Hua,
> I'm afraid I don't understand what you are talking about.
> On Mon, Dec 10, 2018 at 09:28:09AM +0800, Su Hua wrote:
> > Excuse me, the latest drbd driver
> Latest 8.4, or latest 9.0?
> > can provide the arbitration service
> What is the "arbitration service" for DRBD?
> > instead of using the script
> Which script?
> > that is fixed afterwards,
> Who fixed which script in which way and why?
> > which may cause partial data loss under the GFS2 cluster file system.
> About what concrete scenario are you talking?
> Regards, rck
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20181210/ef02cfc8/attachment-0001.htm>

More information about the drbd-user mailing list