|author||Raghavendra G <email@example.com>||2012-12-18 12:47:43 +0530|
|committer||Vijay Bellur <firstname.lastname@example.org>||2013-05-17 00:18:36 -0700|
rpc-transport/rdma: use rdma-cm for connection establishment.
Till now gluster used tcp/ip based communication channel with gluster specific protocol to exchange infiniband addresses. Change-Id: I9de4db398a0e2af51d3d2d68c2fe42168102b190 BUG: 765051 Signed-off-by: Raghavendra G <email@example.com> Reviewed-on: http://review.gluster.org/149 Tested-by: Gluster Build System <firstname.lastname@example.org> Reviewed-by: Vijay Bellur <email@example.com>
Diffstat (limited to 'doc')
1 files changed, 9 insertions, 0 deletions
diff --git a/doc/rdma-cm-in-3.4.0.txt b/doc/rdma-cm-in-3.4.0.txt
new file mode 100644
@@ -0,0 +1,9 @@
+Following is the impact of http://review.gluster.org/#change,149.
+New userspace packages needed:
+rdmacm needs an IPoIB address for connection establishment. This requirement results in following issues:
+* Because of bug #890502, we've to probe the peer on an IPoIB address. This imposes a restriction that all volumes created in the future have to communicate over IPoIB address (irrespective of whether they use gluster's tcp or rdma transport).
+* Currently client has an independence to choose b/w tcp and rdma transports while communicating with the server (by creating volumes with transport-type tcp,rdma). This independence was a byproduct of our ability use the normal channel used with transport-type tcp for rdma connectiion establishment handshake too. However, with new requirement of IPoIB address for connection establishment, we loose this independence (till we bring in multi-network support - where a brick can be identified by a set of ip-addresses and we can choose different pairs of ip-addresses for communication based on our requirements - in glusterd).