Since nobody responded to this, I'd just like to say that the upcoming feature set looks very interesting, and that I'd like to hear other people's experiences with it, and ideas of how to manage multiple peers.<div class="gmail_extra">
<br><br><div class="gmail_quote">On Mon, Dec 3, 2012 at 5:50 AM, Philipp Reisner <span dir="ltr"><<a href="mailto:philipp.reisner@linbit.com" target="_blank">philipp.reisner@linbit.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi,<br>
<br>
After 18 month of work we are happy that we finished the groundwork<br>
for the next generation of DRBD.<br>
<br>
In this development effort we moved all connection related attributes<br>
from the resource object to a dedicated connection object.<br>
A resource can have multiple connections from now on. The advantages<br>
of multiple connections over stacking are:<br>
* IO is tracked by one activity log (with stacking two ALs)<br>
* scales to 31 peers (with stacking to two peers)<br>
* Improved handling<br>
<br>
The (optional) automatic promote feature opens new ways to integrate<br>
DRBD with other cluster managers than pacemaker. It is also<br>
useful to keep the dual-primary time short for live-migrations of<br>
virtual machines in pacemaker clusters.<br>
<br>
Last not least the improvements in the activity log code allow unseen<br>
number of IOPS for DRBD replicated storage if the storage is larger<br>
than the area covered by the AL.<br>
<br>
Please keep in mind that this is the first pre-release, it will be<br>
easy to find and trigger bugs. We are aware of that and will release<br>
improved follow releases regularly. -- Please do not send bug reports for<br>
this pre1 release.<br>
<br>
9.0.0pre1 (api:genl1/proto:86-110)<br>
--------<br>
* Support for multiple connections in a single resource; this feature<br>
obsoletes device stacking; new configuration keywords: connection,<br>
host, node-id, connection-mesh, hosts.<br>
* Automatic promote; if a process opens a drbd device for r/w, drbd<br>
tries to promotes that resource to primary; it the last process<br>
closes the device drbd demotes that resource to secondary; opens<br>
fails if promotion fails. This feature obsoletes become-primary-on<br>
* Activity log striping; Besides the striping this allows AL<br>
sizes of up to 65536.<br>
* Non blocking queuing of AL-updates; This change significantly<br>
improves the number of IOPs in case the workload does not fit into<br>
the configured AL size.<br>
* Resync extents are now 128MiByte instead of 16MiBytes<br>
<br>
Documentation:<br>
<a href="http://www.drbd.org/users-guide-9.0/" target="_blank">http://www.drbd.org/users-guide-9.0/</a><br>
<br>
Code:<br>
<a href="http://oss.linbit.com/drbd/9.0/drbd-9.0.0pre1.tar.gz" target="_blank">http://oss.linbit.com/drbd/9.0/drbd-9.0.0pre1.tar.gz</a><br>
<a href="http://git.drbd.org/gitweb.cgi?p=drbd-9.0.git;a=tag;h=refs/tags/drbd-9.0.0pre1" target="_blank">http://git.drbd.org/gitweb.cgi?p=drbd-9.0.git;a=tag;h=refs/tags/drbd-9.0.0pre1</a><br>
<br>
Best,<br>
Phil<br>
<br>
_______________________________________________<br>
drbd-user mailing list<br>
<a href="mailto:drbd-user@lists.linbit.com">drbd-user@lists.linbit.com</a><br>
<a href="http://lists.linbit.com/mailman/listinfo/drbd-user" target="_blank">http://lists.linbit.com/mailman/listinfo/drbd-user</a><br>
</blockquote></div><br></div>