Whamcloud - gitweb
LU-5570 lnet: check router aliveness timestamp 48/11748/9
authorLiang Zhen <liang.zhen@intel.com>
Thu, 11 Dec 2014 20:53:31 +0000 (15:53 -0500)
committerOleg Drokin <oleg.drokin@intel.com>
Sun, 4 Jan 2015 18:33:03 +0000 (18:33 +0000)
commit339c7b2b784a528f41c432e9b90285d3445b7536
tree3b67a1acc81eceebba217eb9c9f659985e2bcd3d
parentd68d56c00e7788a445f94b6e099f7f8b9e259bae
LU-5570 lnet: check router aliveness timestamp

In current LNet, peer aliveness timestamp is only recorded for peers
on routers, this patch will change this and also record aliveness of
routers on regular nodes, so lnet can check aliveness timestamp of
routers before sending message, avoid to choose router which has old
aliveness timestamp and can be potentially dead or congested.

Another change in this patch is, when lnet got a remote message from
router, it will update remote NI status of router, in this case, lnet
can get up-to-date aliveness informaiton without waiting to the next
RC ping.

Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: I305ba684ced231a87656458382966aeb8ff7ba48
Reviewed-on: http://review.whamcloud.com/11748
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
lnet/include/lnet/lib-lnet.h
lnet/include/lnet/lib-types.h
lnet/lnet/lib-move.c
lnet/lnet/lib-msg.c
lnet/lnet/peer.c
lnet/lnet/router.c
lnet/lnet/router_proc.c