I had tested defining clients by their infiniband address names, but traffic still ended up back to the storage demon over ethernet. But i figured out a work around by setting the storage demon to be normally available on the infiniband address only, in the storage part of the bacula-dir.conf. And then setting forced routes on the few servers that didn't have infiniband, to force them back over ethernet. Seems to be working so far.
My netstat: [r...@is1 ~]# netstat -rn Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 192.168.64.0 0.0.0.0 255.255.255.0 U 0 0 0 ib0 10.0.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 192.239.84.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1 169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 ib0 0.0.0.0 192.239.84.1 0.0.0.0 UG 0 0 0 eth1 [r...@is1 ~]# On Mar 24, 2010, at 7:30 AM, Phil Stracchino wrote: > On 03/24/10 05:56, Thomas Wakefield wrote: >> Phil- >> Your exactly right in what you are thinking. Your diagram is correct, >> the server has both connections. Both interfaces run on separate subnets. >> >> But right now, all the traffic runs over ethernet. Which i have proven >> with a netstat, only the ethernet shows any connections, no connections >> are done over infiniband. > > In that case, I would believe that you have a routing problem, and the > route to the ethernet segment has been assigned a lower (better) routing > metric than the infiniband subnet. What is the output, on your Bacula > server, of 'netstat -rn'? > > You *are* of course defining infiniband clients by their addresses on > the infiniband subnet, right? > > -- > Phil Stracchino, CDK#2 DoD#299792458 ICBM: 43.5607, -71.355 > ala...@caerllewys.net ala...@metrocast.net p...@co.ordinate.org > Renaissance Man, Unix ronin, Perl hacker, Free Stater > It's not the years, it's the mileage. ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
