From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 15422 invoked by alias); 19 Aug 2008 21:20:09 -0000 Received: (qmail 15416 invoked by alias); 19 Aug 2008 21:20:08 -0000 X-Spam-Status: No, hits=-1.7 required=5.0 tests=AWL,BAYES_00,KAM_MX,SPF_HELO_PASS X-Spam-Check-By: sourceware.org X-Spam-Checker-Version: SpamAssassin 3.2.4 (2008-01-01) on bastion.fedora.phx.redhat.com X-Spam-Level: Subject: master - fenced: kill the cluster on misbehaving nodes To: cluster-cvs-relay@redhat.com X-Project: Cluster Project X-Git-Module: cluster.git X-Git-Refname: refs/heads/master X-Git-Reftype: branch X-Git-Oldrev: 34679f895b267f6efdfa83a63aa65d09e631d5ed X-Git-Newrev: 55e768917ecd4e3e252d5606c10f9cf0dda54d94 From: David Teigland Message-Id: <20080819211858.F41661202A5@lists.fedorahosted.org> Date: Tue, 19 Aug 2008 21:28:00 -0000 X-Scanned-By: MIMEDefang 2.58 on 172.16.52.254 Mailing-List: contact cluster-cvs-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Post: List-Help: , Sender: cluster-cvs-owner@sourceware.org X-SW-Source: 2008-q3/txt/msg00283.txt.bz2 Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=55e768917ecd4e3e252d5606c10f9cf0dda54d94 Commit: 55e768917ecd4e3e252d5606c10f9cf0dda54d94 Parent: 34679f895b267f6efdfa83a63aa65d09e631d5ed Author: David Teigland AuthorDate: Tue Aug 19 16:09:18 2008 -0500 Committer: David Teigland CommitterDate: Tue Aug 19 16:09:18 2008 -0500 fenced: kill the cluster on misbehaving nodes Kill cman on other nodes where the fenced process fails. Signed-off-by: David Teigland --- fence/fenced/cpg.c | 3 +++ fence/fenced/fd.h | 1 + fence/fenced/member_cman.c | 22 ++++++++++++++++++++-- 3 files changed, 24 insertions(+), 2 deletions(-) diff --git a/fence/fenced/cpg.c b/fence/fenced/cpg.c index 1a0ff17..ccebbd9 100644 --- a/fence/fenced/cpg.c +++ b/fence/fenced/cpg.c @@ -1069,6 +1069,9 @@ static int add_change(struct fd *fd, log_debug("add_change %u nodeid %d remove reason %d", cg->seq, memb->nodeid, left_list[i].reason); + + if (left_list[i].reason == CPG_REASON_PROCDOWN) + kick_node_from_cluster(memb->nodeid); } for (i = 0; i < joined_list_entries; i++) { diff --git a/fence/fenced/fd.h b/fence/fenced/fd.h index 96c8a1d..21cac39 100644 --- a/fence/fenced/fd.h +++ b/fence/fenced/fd.h @@ -249,6 +249,7 @@ int is_cman_member(int nodeid); char *nodeid_to_name(int nodeid); int name_to_nodeid(char *name); struct node *get_new_node(struct fd *fd, int nodeid); +void kick_node_from_cluster(int nodeid); /* recover.c */ diff --git a/fence/fenced/member_cman.c b/fence/fenced/member_cman.c index 7820d64..16e08a8 100644 --- a/fence/fenced/member_cman.c +++ b/fence/fenced/member_cman.c @@ -5,9 +5,21 @@ #define BUFLEN MAX_NODENAME_LEN+1 static cman_handle_t ch; +static cman_handle_t ch_admin; static cman_node_t cman_nodes[MAX_NODES]; static int cman_node_count; +void kick_node_from_cluster(int nodeid) +{ + if (!nodeid) { + log_error("telling cman to shut down cluster locally"); + cman_shutdown(ch_admin, CMAN_SHUTDOWN_ANYWAY); + } else { + log_error("telling cman to remove nodeid %d from cluster", + nodeid); + cman_kill_node(ch_admin, nodeid); + } +} static int name_equal(char *name1, char *name2) { @@ -146,12 +158,18 @@ int setup_cman(void) int init = 0, active = 0; retry_init: - ch = cman_init(NULL); - if (!ch) { + ch_admin = cman_admin_init(NULL); + if (!ch_admin) { if (init++ < 2) { sleep(1); goto retry_init; } + log_error("cman_admin_init error %d", errno); + return -ENOTCONN; + } + + ch = cman_init(NULL); + if (!ch) { log_error("cman_init error %d", errno); return -ENOTCONN; }