From: root Date: Mon, 30 Jul 2018 01:29:48 +0000 (-0400) Subject: msg: ceph_abort() when there are enough accepter errors in msg server X-Git-Tag: v13.2.3~20^2 X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=bfe74e9a8545a852eda7aee4d41dec79e0bfa10d;p=ceph.git msg: ceph_abort() when there are enough accepter errors in msg server In some extrem cases(we have met one in our production cluster), when Accepter thread break out , new client can not connect to the osd. Because the former heartbeat connections are already connected, other osd can not detect failure then notify monitor to mark the failed osd down. In the patch, we there are abnormal communication errors ,we just ceph_abort so that osd can go down fastly and other osds can notify monitor to mark the failed osd down. Signed-off-by: penglaiyxy@gmail.com (cherry picked from commit 00e0ab407b2e9659d9121be1217e95c8117c411e) Conflicts: src/common/legacy_config_opts.h src/common/options.cc src/msg/async/AsyncMessenger.cc: trivial resolution --- diff --git a/src/common/legacy_config_opts.h b/src/common/legacy_config_opts.h index b76e089732dc..e2bb67fb13f0 100644 --- a/src/common/legacy_config_opts.h +++ b/src/common/legacy_config_opts.h @@ -166,6 +166,10 @@ OPTION(ms_async_rdma_roce_ver, OPT_INT) // 0=RoCEv1, 1=RoCEv2, 2=RoCEv1. OPTION(ms_async_rdma_sl, OPT_INT) // in RoCE, this means PCP OPTION(ms_async_rdma_dscp, OPT_INT) // in RoCE, this means DSCP +// when there are enough accept failures, indicating there are unrecoverable failures, +// just do ceph_abort() . Here we make it configurable. +OPTION(ms_max_accept_failures, OPT_INT) + OPTION(ms_dpdk_port_id, OPT_INT) SAFE_OPTION(ms_dpdk_coremask, OPT_STR) // it is modified in unittest so that use SAFE_OPTION to declare OPTION(ms_dpdk_memory_channel, OPT_STR) diff --git a/src/common/options.cc b/src/common/options.cc index 6857b8aa0940..197e44daab95 100644 --- a/src/common/options.cc +++ b/src/common/options.cc @@ -1044,6 +1044,11 @@ std::vector