[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [dm-devel] RE: How is the multipath pg policy group-by-nodename any differen t than the failover policy?

On 28 Apr 2005, at 19:53, goggin, edward wrote:

On Wed, 27 Apr 2005 21:38:34 +0100, Arthur Bergman wrote:

All the path have the same serial, but the show up from different nodes.

Not sure what this means. The group-by-node-name policy implies assigning
paths with similar node name attribute to the same group. For SCSI, seems
like the multipath configurator assigns the node name attribute to be the
combination of SCSI initiator and target addresses (in linux, host
no/channel no/target id). This should be different for every path which
would cause each path to be assigned to a different path group.

How is the resultant path grouping any different than if you used the
path grouping policy instead?

dm-devel mailing list
dm-devel redhat com

Huh, as far as I understand it group_by_nodename uses the

/sys/class/fc_transport/target0:0:0/node_name information

from our setup, there are 8 paths to two controllers, there is no failover cost but it defeats read caching pretty beadly to use both controllers..


So all path that ends at a matching node_name, even if they are on different transport targets get one path.


CTO @ Fotango Ltd

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]