[ClusterLabs] Teamed interfaces and Corosync v2

Digimer lists at alteeve.ca
Sat Apr 8 20:30:06 EDT 2017

Hi all,

  I know that under corosync v1 and traditional bonding, only a subset
of modes were supported (by Red Hat, anyway). Mode 1 initially, then I
think 0 and 2 later.

  I'm looking at moving to EL7 (finally!) and I notice that teamed
interfaces are replacing bonded interfaces, and a new mode caught my
attention; Broadcast;

broadcast (data is transmitted over all ports)

active-backup (one port or link is used while others are kept as a backup)

  If I am understanding this right, using broadcast would be even faster
failover than active-backup, as there would be no actual failover. If I
am interpreting this correctly, then it seems to me that I would want to
use this under corosync traffic. Note that I am totally new to teamd and
might be making a dumb assumption here. :)

  So questions, two;

1. Is broadcast more resilient than active-backup?

2. Are there any contraindications for using broadcast teams (or teamd
at all) under corosync v2?


Papers and Projects: https://alteeve.com/w/
"I am, somehow, less interested in the weight and convolutions of
Einstein’s brain than in the near certainty that people of equal talent
have lived and died in cotton fields and sweatshops." - Stephen Jay Gould

More information about the Users mailing list