cluster-commits April 2011

cluster-commits@lists.fedorahosted.org

4 participants
37 discussions

cluster: RHEL6 - fence_ipmilan: Correct return code for diag operation
by Lon Hohberger 06 Apr '11

06 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=e5… Commit: e51fbc090cb3b7a343aae3ff7d1ffcd94f7fbb5b Parent: 3ca792d705ca4c52a7224c183eaa84b38682557f Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 15:19:57 2011 +0200 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Wed Apr 6 09:47:24 2011 -0400 fence_ipmilan: Correct return code for diag operation Return code for diag operation should be 0. Replaces commit 8a97abd04b8492310c1095413c5cf26bbf98e395 "diaf" != "diag" Resolves: rhbz#655764 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> Tested-by: Lon Hohberger <lhh(a)redhat.com> --- fence/agents/ipmilan/ipmilan.c | 10 ++++++++++ 1 files changed, 10 insertions(+), 0 deletions(-) diff --git a/fence/agents/ipmilan/ipmilan.c b/fence/agents/ipmilan/ipmilan.c index be36f58..92009f0 100644 --- a/fence/agents/ipmilan/ipmilan.c +++ b/fence/agents/ipmilan/ipmilan.c @@ -1189,5 +1189,15 @@ metaout: else printf("Failed\n"); } + + if (!strcasecmp(op, "diag")) { + /** .. but when a system receive the DIAG signal , it + ** switches to a kdump kernel, but the machine is always + ** "on" during the dump phase. It only become temporarily + ** "off" at the end of the dump just before rebooting. + **/ + translated_ret = 0; + } + return translated_ret; }

1 0

cluster: STABLE31 - rgmanager: re-add slang scripts where they belong
by Fabio M. Di Nitto 05 Apr '11

05 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=62… Commit: 62cc660f462f8672fe75ebc8e53b184075ff3d5b Parent: d8a4abe7ff8ff0954fb58fed9b384b087e81b1d0 Author: Fabio M. Di Nitto <fdinitto(a)redhat.com> AuthorDate: Tue Apr 5 10:35:48 2011 +0200 Committer: Fabio M. Di Nitto <fdinitto(a)redhat.com> CommitterDate: Tue Apr 5 10:35:48 2011 +0200 rgmanager: re-add slang scripts where they belong Signed-off-by: Fabio M. Di Nitto <fdinitto(a)redhat.com> --- rgmanager/src/Makefile | 2 +- rgmanager/src/resources/Makefile | 12 + rgmanager/src/resources/default_event_script.sl | 637 +++++++++++++++++++++++ rgmanager/src/resources/follow-service.sl | 151 ++++++ 4 files changed, 801 insertions(+), 1 deletions(-) diff --git a/rgmanager/src/Makefile b/rgmanager/src/Makefile index f5b9d03..c79d053 100644 --- a/rgmanager/src/Makefile +++ b/rgmanager/src/Makefile @@ -1,4 +1,4 @@ include ../../make/defines.mk include $(OBJDIR)/make/passthrough.mk -SUBDIRS= clulib daemons utils +SUBDIRS=resources clulib daemons utils diff --git a/rgmanager/src/resources/Makefile b/rgmanager/src/resources/Makefile new file mode 100644 index 0000000..00b2bc8 --- /dev/null +++ b/rgmanager/src/resources/Makefile @@ -0,0 +1,12 @@ +include ../../../make/defines.mk + +all: + +SHAREDIRT= default_event_script.sl \ + follow-service.sl + +include $(OBJDIR)/make/clean.mk +include $(OBJDIR)/make/install.mk +include $(OBJDIR)/make/uninstall.mk + +clean: generalclean diff --git a/rgmanager/src/resources/default_event_script.sl b/rgmanager/src/resources/default_event_script.sl new file mode 100644 index 0000000..87c8a1c --- /dev/null +++ b/rgmanager/src/resources/default_event_script.sl @@ -0,0 +1,637 @@ +% +% Copyright (C) 1997-2003 Sistina Software, Inc. All rights reserved. +% Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved. +% +% This program is free software; you can redistribute it and/or +% modify it under the terms of the GNU General Public License +% as published by the Free Software Foundation; either version 2 +% of the License, or (at your option) any later version. +% +% This program is distributed in the hope that it will be useful, +% but WITHOUT ANY WARRANTY; without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +% GNU General Public License for more details. +% +% You should have received a copy of the GNU General Public License +% along with this program; if not, write to the Free Software +% Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. +% + + +define node_in_set(node_list, node) +{ + variable x, len; + + len = length(node_list); + for (x = 0; x < len; x++) { + if (node_list[x] == node) + return 1; + } + + return 0; +} + + +% +% Returns 3 node lists: +% (1) Nodes with no services +% (2) Nodes with non-exclusive services +% (3) Nodes with exclusive services +% +% NOTE: This function currently defenstrates failover domain rules +% +define separate_nodes(node_list) +{ + variable services = service_list(); + variable nodes_empty, nodes_services, nodes_excl; + variable x, len; + variable owner, state, excl, ns = 0, nx = 0; + + nodes_empty = node_list; + + % Most Awesome Initializer EVER!!! + nodes_services = subtract([0], 0); + nodes_excl = subtract([0], 0); + + len = length(services); + for (x = 0; x < len; x++) { + + (,,, owner, state) = service_status(services[x]); + if (owner < 0) { + continue; + } + + excl = atoi(service_property(services[x], "exclusive")); + nodes_empty = subtract(nodes_empty, owner); + if (excl) { + nodes_excl = union(nodes_excl, owner); + } else { + nodes_services = union(nodes_services, owner); + } + } + + return (nodes_empty, nodes_services, nodes_excl); +} + + +define exclusive_prioritize(svc, node_list) +{ + variable services = service_list(); + variable len, x, y, owner, state, preferred_owner; + variable svc_excl, other_excl; + variable nodes_x, nodes_s, nodes_e; + + % + % Not exclusive? Don't care! + % + svc_excl = atoi(service_property(svc, "exclusive")); + if (svc_excl == 0) { + return node_list; + } + + (nodes_e, nodes_s, nodes_x) = separate_nodes(node_list); + debug("Nodes - Empty: ", nodes_e, " w/Services: ", nodes_s, " w/Excl: ", nodes_x); + if (length(nodes_e) > 0) { + % + % If we've got an exclusive service, only allow it to start on + % empty nodes. + % + return nodes_e; + } + + if (length(nodes_x) == 0) { + % + % If we've got NO nodes with other exclusive services + % and no empty nodes, the service can not be started + % + notice("No empty / exclusive nodes available; cannot restart ", svc); + return nodes_x; + } + + % + % Prioritization of exclusive services: pancake a service and replace it + % with this service if this services is a higher priority. + % + len = length(services); + for (x = 0; x < len; x++) { + if (svc == services[x]) { + % don't do anything to ourself! + continue; + } + + (,,, owner, state) = service_status(services[x]); + if (owner < 0) { + continue; + } + + if (node_in_set(node_list, owner) == 0) { + continue; + } + + other_excl = atoi(service_property(services[x], "exclusive")); + if (other_excl == 0) { + continue; + } + + % + % If we're a higher priority (lower #) exclusive + % Stop the exclusive service that node and move that + % node to the front. + % + if (svc_excl >= other_excl) { + continue; + } + + % + % + % + warning("STOPPING service ", services[x], " because ", svc, " is a higher priority."); + () = service_stop(services[x]); + + % + % Return just the one node. + % + node_list = subtract([0], 0); + node_list = union(node_list, owner); + return node_list; + } + + return node_list; +} + + +define move_or_start(service, node_list) +{ + variable len; + variable state, owner; + variable depends; + + depends = service_property(service, "depend"); + if (depends != "") { + (,,, owner, state) = service_status(depends); + if (owner < 0) { + debug(service, " is not runnable; dependency not met"); + return ERR_DEPEND; + } + } + + (,,, owner, state) = service_status(service); + debug("Evaluating ", service, " state=", state, " owner=", owner); + if ((event_type == EVENT_NODE) and (node_id == owner) and + (node_state == NODE_OFFLINE)) { + info("Marking service ", service, " on down member ", + owner, " as stopped"); + if (service_stop(service) < 0) { + return ERR_ABORT; + } + } + + len = length(node_list); + if (len == 0) { + notice(service, " is not runnable - restricted domain offline"); + ()=service_stop(service); + return ERR_DOMAIN; + } + + if (((event_type != EVENT_USER) and (state == "disabled")) or + ((state == "failed") or (state == "frozen"))) { + % + % Commenting out this block will -not- allow you to + % recover failed services from event scripts. Sorry. + % All it will get you is a false log message about + % starting this service. + % + % You may enable disabled services, but I recommend + % against it. + % + debug(service, " is not runnable"); + return -1; + } + + if (node_list[0] == owner) { + debug(service, " is already running on best node"); + return ERR_RUNNING; + } + + if ((owner >= 0) and (node_in_set(node_list, owner) == 1)) { + notice("Moving ", service, " from ", owner, + " to ", node_list); + if (service_stop(service) < 0) { + return ERR_ABORT; + } + } else { + node_list = exclusive_prioritize(service, node_list); + notice("Starting ", service, " on ", node_list); + } + + if (length(node_list) == 0) { + return ERR_DOMAIN; + } + return service_start(service, node_list); +} + + +% +% Returns the set of online nodes in preferred/shuffled order which +% are allowed to run this service. Gives highest preference to current +% owner if nofailback is specified. +% +define allowed_nodes(service) +{ + variable anodes; + variable online; + variable nodes_domain; + variable ordered, restricted, nofailback; + variable state, owner; + variable depends; + + (nofailback, restricted, ordered, nodes_domain) = + service_domain_info(service); + + (,,, owner, state) = service_status(service); + + anodes = nodes_online(); + + % Shuffle the array so we don't start all services on the same + % node. TODO - add RR, Least-services, placement policies... + online = shuffle(anodes); + + if (restricted == 1) { + anodes = intersection(nodes_domain, online); + } else { + % Ordered failover domains (nodes_domain) unioned with the + % online nodes basically just reorders the online node list + % according to failover domain priority rules. + anodes = union(intersection(nodes_domain, online), + online); + } + + if ((nofailback == 1) or (ordered == 0)) { + + if ((owner < 0) or (node_in_set(anodes, owner) == 0)) { + return anodes; + } + + % Because union takes left as priority, we can + % return the union of the current owner with the + % allowed node list. This means the service will + % remain on the same node it's currently on. + return union(owner, anodes); + } + + return anodes; +} + +% +% Returns the set of online nodes in preferred/shuffled order which +% are allowed to run this service. Gives highest preference to current +% owner if nofailback is specified. +% +define allowed_nodes(service) +{ + variable anodes; + variable online; + variable nodes_domain; + variable ordered, restricted, nofailback; + variable state, owner; + variable depends; + + (nofailback, restricted, ordered, nodes_domain) = + service_domain_info(service); + (,,, owner, state) = service_status(service); + + anodes = nodes_online(); + + % Shuffle the array so we don't start all services on the same + % node. TODO - add RR, Least-services, placement policies... + online = shuffle(anodes); + + if (restricted == 1) { + anodes = intersection(nodes_domain, online); + } else { + % Ordered failover domains (nodes_domain) unioned with the + % online nodes basically just reorders the online node list + % according to failover domain priority rules. + anodes = union(intersection(nodes_domain, online), + online); + } + + if ((nofailback == 1) or (ordered == 0)) { + + if ((owner < 0) or (node_in_set(anodes, owner) == 0)) { + return anodes; + } + + % Because union takes left as priority, we can + % return the union of the current owner with the + % allowed node list. This means the service will + % remain on the same node it's currently on. + return union(owner, anodes); + } + + return anodes; +} + +define string_list(thelist, delimiter) +{ + variable index; + variable output=""; + + if (length(thelist) == 0) { + return output; + } + + for (index=0; index < length(thelist)-1; index++) { + output=output+string(thelist[index])+delimiter; + } + return output+thelist[index]; +} + +% this function gets the smallest property from a given list of services +% if the list only exists of one element the property itself is returned +% if the given property is not found 0 is returned +define services_min_attribute(services, property) +{ + variable x; + variable min_property=-1; + variable tmp_property; + + for (x = 0; x < length(services); x++) { + tmp_property=service_property(services[x], property); + if (tmp_property == NULL) { + tmp_property=0; + } else { + tmp_property=atoi(tmp_property); + } + if ((min_property < 0) or (tmp_property < min_property)) { + min_property=tmp_property; + } + %debug("services_min_attribute: ",services[x]," attribute: ",min_property, "tmp: ", tmp_property, " min: ", min_property); + } + + %debug("services_min_attribute: (", string_list(services, ", "),")[",property,"]: ",min_property); + + return min_property; +} + +% This function will sort a given service_list by the given attribute name and +% return the list +define sorted_service_list(services, attribute) +{ + variable work_queue={}; + variable sorted_list={}, tmp, tmp2; + variable x, y; + variable cur_min_prop=0; + variable service_prop=0; + + y=0; + %debug("sorted_service_list: ", strjoin(services, ", ")); + for (x=0; x<length(services); x++) { + list_append(work_queue, string(services[x])); + } + + %debug("sorted_service_list: work_queue ", string_list(work_queue, ", ")); + while (length(work_queue) > 0) { + cur_min_prop=services_min_attribute(work_queue, attribute); + %debug("sorted_service_list sorting services list for attribute ", attribute, " cur_min: ",cur_min_prop); + for (x = 0; x < length(work_queue); x++) { + service_prop=service_property(work_queue[x], "priority"); + if (service_prop == NULL) { + service_prop=0; + } else { + service_prop=atoi(service_prop); + } + %debug("sorted_service_list: ",work_queue[x], " property[", attribute,"]: ",service_prop); + if (cur_min_prop==service_prop) { + %debug("sorted_service_list: adding service ",work_queue[x]," to sorted. work_queue: ", string_list(work_queue, ", ")); + list_append(sorted_list, work_queue[x]); + %debug("sorted_service_list: sorted_list: ", string_list(sorted_list, ", ")); + %debug("sorted_service_list: removing service ",work_queue[x], " from work_queue ", string_list(work_queue, ", ")); + list_delete(work_queue, x); + x=x-1; + %debug("sorted_service_list: work_queue: ",string_list(work_queue, ", ")); + y=y+1; + } + } + } + + debug("sorted_service_list ", string_list(sorted_list, ", ")); + return sorted_list; +} + +define sortedservices_node_event_handler(services, attribute) { + variable x; + variable nodes; + + services=sorted_service_list(services, attribute); + for (x = 0; x < length(services); x++) { + debug("Executing sortedservices node event handler for service: ", services[x]); + nodes = allowed_nodes(services[x]); + ()=move_or_start(services[x], nodes); + } +} + +define default_node_event_handler() +{ + variable services = service_list(); + variable x; + variable nodes; + + debug("Executing default node event handler"); + for (x = 0; x < length(services); x++) { + nodes = allowed_nodes(services[x]); + ()=move_or_start(services[x], nodes); + } +} + + +define default_service_event_handler() +{ + variable services = service_list(); + variable x; + variable depends; + variable depend_mode; + variable policy; + variable nodes; + variable tmp; + variable owner; + variable state; + + debug("Executing default service event handler"); + + if (service_state == "recovering") { + + policy = service_property(service_name, "recovery"); + debug("Recovering", + " Service: ", service_name, + " Last owner: ", service_last_owner, + " Policy: ", policy, + " RTE: ", service_restarts_exceeded); + + if (policy == "disable") { + () = service_stop(service_name, 1); + return; + } + + nodes = allowed_nodes(service_name); + if (policy == "restart" and service_restarts_exceeded == 0) { + nodes = union(service_last_owner, nodes); + } else { + % relocate + tmp = subtract(nodes, service_last_owner); + if (length(tmp) == 0) { + () = service_stop(service_name,0); + return; + } + + nodes = union(tmp, service_last_owner); + } + + ()=move_or_start(service_name, nodes); + + return; + } + + for (x = 0; x < length(services); x++) { + if (service_name == services[x]) { + % don't do anything to ourself! + continue; + } + + % + % Simplistic dependency handling + % + depends = service_property(services[x], "depend"); + depend_mode = service_property(services[x], "depend_mode"); + + % No dependency; do nothing + if (depends != service_name) { + continue; + } + + (,,, owner, state) = service_status(services[x]); + if ((service_state == "started") and (owner < 0) and + (state == "stopped")) { + info("Dependency met; starting ", services[x]); + nodes = allowed_nodes(services[x]); + ()=move_or_start(services[x], nodes); + } + + % service died - stop service(s) that depend on the dead + if ((service_owner < 0) and (owner >= 0) and + (depend_mode != "soft")) { + info("Dependency lost; stopping ", services[x]); + ()=service_stop(services[x]); + } + } +} + +define default_config_event_handler() +{ + debug("Executing default config event handler"); +} + +define default_user_event_handler() +{ + variable ret; + variable nodes; + variable reordered; + variable x; + variable target = user_target; + variable found = 0; + variable owner, state; + + nodes = allowed_nodes(service_name); + (,,, owner, state) = service_status(service_name); + + if (user_request == USER_RESTART) { + + if (owner >= 0) { + reordered = union(owner, nodes); + nodes = reordered; + } + + notice("Stopping ", service_name, " for relocate to ", nodes); + + found = service_stop(service_name); + if (found < 0) { + return ERR_ABORT; + } + + ret = move_or_start(service_name, nodes); + + } else if ((user_request == USER_RELOCATE) or + (user_request == USER_ENABLE)) { + + if (user_target > 0) { + for (x = 0; x < length(nodes); x++) { + % + % Put the preferred node at the front of the + % list for a user-relocate operation + % + if (nodes[x] == user_target) { + reordered = union(user_target, nodes); + nodes = reordered; + found = 1; + } + } + + if (found == 0) { + warning("User specified node ", user_target, + " is offline"); + } + } + + if ((owner >= 0) and (user_request == USER_RELOCATE)) { + if (service_stop(service_name) < 0) { + return ERR_ABORT; + } + + % + % The current owner shouldn't be the default + % for a relocate operation + % + reordered = subtract(nodes, owner); + nodes = union(reordered, owner); + } + + ret = move_or_start(service_name, nodes); + + } else if (user_request == USER_DISABLE) { + + ret = service_stop(service_name, 1); + + } else if (user_request == USER_STOP) { + + ret = service_stop(service_name); + + } else if (user_request == USER_FREEZE) { + + ret = service_freeze(service_name); + + } else if (user_request == USER_UNFREEZE) { + + ret = service_unfreeze(service_name); + + } else if (user_request == USER_MIGRATE) { + + ret = service_migrate(service_name, user_target); + + } else if (user_request == USER_CONVALESCE) { + + ret = service_convalesce(service_name); + + } + + return ret; +} + +if (event_type == EVENT_NODE) + sortedservices_node_event_handler(service_list(), "priority"); +if (event_type == EVENT_SERVICE) + default_service_event_handler(); +if (event_type == EVENT_CONFIG) + default_config_event_handler(); +if (event_type == EVENT_USER) + user_return=default_user_event_handler(); + diff --git a/rgmanager/src/resources/follow-service.sl b/rgmanager/src/resources/follow-service.sl new file mode 100644 index 0000000..4c711ec --- /dev/null +++ b/rgmanager/src/resources/follow-service.sl @@ -0,0 +1,151 @@ +% follow-service.sl +% +% Description: Implements the "follow service" mechanism based on the Red Hat RIND event +% scripting mechanism. +% +% Author: Marc Grimme, Mark Hlawatschek, October 2008 +% Support: support(a)atix.de +% License: GNU General Public License (GPL), version 2 or later +% Copyright: (c) 2008-2010 ATIX AG + + +debug("*** follow-service.sl"); + + +% +% Returns a list of nodes for the given service that are online and in the failoverdomain. +% +define nodelist_online(service_name) { + variable nodes, nofailback, restricted, ordered, node_list; + nodes=nodes_online(); + + (nofailback, restricted, ordered, node_list) = service_domain_info(service_name); + + return intersection(nodes, node_list); +} + +% +% Idea: +% General purpose function of a construct when Service(svc1) and Service(svc2) +% should not be running on the same node even after failover. +% There are to options to influence the behaviour. If both services have to be +% running on the same node (only one node is left in the failovergroup) what +% service is the master and should both services be running or only the master +% service survives. If master is not svc1 or svc2 both service might run on the +% same node. If master is either svc1 or svc2 the specified one will be the +% surviving service. +% If followslave is not 0 the svc1 always follows svc2. That means it will be +% started on on the same node as svc1. And if available svc2 will be relocated +% to any other node. +% +define follow_service(svc1, svc2, master) %, followslave) +{ + variable state_svc1, state_svc2, owner_svc1, owner_svc2; + variable nodes1, nodes2, allowed; + + debug("*** FOLLOW_SERVICE: follow_service(",svc1,", ",svc2,", ", master, ")"); + debug("*** FOLLOW_SERVICE: event_type: ", event_type, ", service_name: ", service_name, ", service_state: ", service_state); + + % + % setup the master + % + if ((master != svc1) and (master != svc2)) { + debug("*** FOLLOW_SERVICE: master=NULL"); + master=NULL; + } + + % get infos we need to decide further + (,,, owner_svc1, state_svc1) = service_status(svc1); + (,,, owner_svc2, state_svc2) = service_status(svc2); + nodes1 = nodelist_online(svc1); + nodes2 = nodelist_online(svc2); + debug("*** FOLLOW_SERVICE: service_status(",svc1,"): ", state_svc1); + debug("*** FOLLOW_SERVICE: owner_svc1: ", owner_svc1, ", owner_svc2: ", owner_svc2, ", nodes1: ", nodes1, ", nodes2: ", nodes2); + + if (((event_type == EVENT_NODE) and (owner_svc1 == node_id) and (node_state == NODE_OFFLINE) and (owner_svc2 >=0)) or + ((event_type == EVENT_SERVICE) and (service_name == svc1) and (service_state == "recovering" ) and (owner_svc2 >= 0))) { + % + % uh oh, the owner of the master server died. Restart it + % on the node running the slave server or if we should not + % follow the slave start it somewhere else. + % We should end up here if svc1 has to be restarted + + % + % If this was a service event, don't execute the default event + % script trigger after this script completes. + % + if (event_type == EVENT_SERVICE) { + stop_processing(); + } + % were to start svc2 + allowed=subtract(nodes2, owner_svc2); + if (length(allowed) > 1) { + allowed=subtract(allowed, service_last_owner); + } + debug("*** FOLLOW SERVICE: service event triggered following svc2 to ",owner_svc2, " svc2 on : ",allowed); + + % either svc1 is the master or there are node were to start svc2 + if ((master == svc1) or (length(allowed) > 0)) { + ()=service_start(svc1, owner_svc2); + } + % either svc2 is the master or there are node were to start svc2 + if ((master == svc2) or (length(allowed) > 0)) { + ()=service_stop(svc2); + ()=service_start(svc2, allowed); + } + } + else if (((event_type == EVENT_NODE) and (owner_svc2 == node_id) and (node_state == NODE_OFFLINE) and (owner_svc2 >=0)) or + ((event_type == EVENT_SERVICE) and (service_name == svc2) and (service_state == "recovering" ) and (owner_svc1 >= 0))) { + % + % uh oh, the owner of the svc2 died. Restart it + % on any other node but not the one running the svc1. + % If svc1 is the only one left only start it there + % if master==svc2 + % + % Just relocate svc2 or if svc2 is master stop svc1 and start svc2 on owner_svc1 + + % + % If this was a service event, don't execute the default event + % script trigger after this script completes. + % + + if (event_type == EVENT_SERVICE) { + stop_processing(); + } + + allowed=subtract(nodes2, owner_svc1); + if (length(allowed) > 1) { + allowed=subtract(allowed, service_last_owner); + } + + debug("*** FOLLOW SERVICE: service event triggered relocating svc2 to ",allowed, " svc1 on : ",owner_svc1); + + if (length(allowed) > 0) { + ()=service_stop(svc2); + ()=service_start(svc2, allowed); + } else if (master == svc2) { + ()=service_stop(svc1); + ()=service_start(svc2, owner_svc1); + } + } + else if (((event_type == EVENT_SERVICE) and (service_state == "started") and (owner_svc2 == owner_svc1) and (owner_svc1 > 0) and (owner_svc2 > 0)) or + ((event_type == EVENT_CONFIG) and (owner_svc2 == owner_svc1))) { + allowed=subtract(nodes2, owner_svc1); + debug("*** FOLLOW SERVICE: service event both running on same node triggered.", allowed); + if (length(allowed) > 0) { + %()=service_stop(svc1); + %()=service_start(svc1, owner_svc2); + ()=service_stop(svc2); + ()=service_start(svc2, allowed); + } else if ((master == svc2) and (owner_svc2 > 0)){ + debug("*** FOLLOW SERVICE: will stop service .", svc1); + ()=service_stop(svc1); + } else if ((master == svc1) and (owner_svc1 > 0)) { + debug("*** FOLLOW SERVICE: will stop service .", svc2); + ()=service_stop(svc2); + } else { + debug("*** FOLLOW SERVICE: both services running on the same node or only one is running.", allowed, ", ", master); + } + } + return; +}

1 0

cluster: STABLE31 - cman: Don't increment on LEAVING->MEMBER transition
by Lon Hohberger 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=d8… Commit: d8a4abe7ff8ff0954fb58fed9b384b087e81b1d0 Parent: e6e7f1837b53d8f2095392afc30aca411307730c Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Mon Dec 6 15:01:55 2010 -0500 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Mon Apr 4 16:58:34 2011 -0400 cman: Don't increment on LEAVING->MEMBER transition If a node left the cluster prior using 'cman_tool leave', its internal state is set to NODESTATE_LEAVING. Ordinarily, during the transition from NODESTATE_LEAVING to NODESTATE_DEAD, the member count is decremented. However, it is possible to rejoin the cluster membership from the NODESTATE_LEAVING state. If this occurs, the cluster member count is incremented, but since the node never transitioned from LEAVING->DEAD, it was not previously decremented, causing an ever-increasing node count in 'cman_tool status'. This patch resolves the issue by not incrementing the member count during a LEAVING->MEMBER transition. Resolves: rhbz#649533 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- cman/daemon/commands.c | 7 +++++-- 1 files changed, 5 insertions(+), 2 deletions(-) diff --git a/cman/daemon/commands.c b/cman/daemon/commands.c index d4d9fd9..2948952 100644 --- a/cman/daemon/commands.c +++ b/cman/daemon/commands.c @@ -2246,9 +2246,12 @@ void add_ais_node(int nodeid, uint64_t incar, int total_members) if (node->state == NODESTATE_DEAD || node->state == NODESTATE_LEAVING) { gettimeofday(&node->join_time, NULL); node->incarnation = incar; - node->state = NODESTATE_MEMBER; node->leave_reason = 0; - cluster_members++; + /* If a node rejoins before it completes a leave, + * we should not increment cluster_members */ + if (node->state != NODESTATE_LEAVING) + cluster_members++; + node->state = NODESTATE_MEMBER; recalculate_quorum(0, 0); } }

1 0

fence-agents: master - fence_ipmilan: Correct return code for diag operation
by Lon Hohberger 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/fence-agents.git?p=fence-agents.git;a=commi… Commit: 550a6ca4caee40bdf669e563e66371b23e564d6e Parent: ed60a64a1eba5c941edaf7802e2d6ab3c638aaa5 Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 15:19:57 2011 +0200 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Mon Apr 4 16:12:28 2011 -0400 fence_ipmilan: Correct return code for diag operation Return code for diag operation should be 0. Replaces commit 8a97abd04b8492310c1095413c5cf26bbf98e395 "diaf" != "diag" Resolves: rhbz#655764 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- fence/agents/ipmilan/ipmilan.c | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-) diff --git a/fence/agents/ipmilan/ipmilan.c b/fence/agents/ipmilan/ipmilan.c index 040babf..892fc4e 100644 --- a/fence/agents/ipmilan/ipmilan.c +++ b/fence/agents/ipmilan/ipmilan.c @@ -1195,6 +1195,15 @@ metaout: if (!strcasecmp(op, "monitor")) translated_ret = ret; + + if (!strcasecmp(op, "diag")) { + /** .. but when a system receive the DIAG signal , it + ** switches to a kdump kernel, but the machine is always + ** "on" during the dump phase. It only become temporarily + ** "off" at the end of the dump just before rebooting. + **/ + translated_ret = 0; + } return translated_ret; }

1 0

fence-agents: master - Revert "fence_ipmilan: Correct return code for diag operation"
by Lon Hohberger 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/fence-agents.git?p=fence-agents.git;a=commi… Commit: ed60a64a1eba5c941edaf7802e2d6ab3c638aaa5 Parent: bf66212512ad7f97b38fcb49fc256ac76583afbd Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Mon Apr 4 16:11:43 2011 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Mon Apr 4 16:11:43 2011 -0400 Revert "fence_ipmilan: Correct return code for diag operation" This reverts commit 8a97abd04b8492310c1095413c5cf26bbf98e395. --- fence/agents/ipmilan/ipmilan.c | 9 --------- 1 files changed, 0 insertions(+), 9 deletions(-) diff --git a/fence/agents/ipmilan/ipmilan.c b/fence/agents/ipmilan/ipmilan.c index 177d216..040babf 100644 --- a/fence/agents/ipmilan/ipmilan.c +++ b/fence/agents/ipmilan/ipmilan.c @@ -1195,15 +1195,6 @@ metaout: if (!strcasecmp(op, "monitor")) translated_ret = ret; - - if (!strcasecmp(op, "diaf")) { - /** .. but when a system receive the DIAG signal , it - ** switches to a kdump kernel, but the machine is always - ** "on" during the dump phase. It only become temporarily - ** "off" at the end of the dump just before rebooting. - **/ - translated_ret = 0; - } return translated_ret; }

1 0

cluster: RHEL57 - fence_vmware_soap: New fence agent for VMWare using SOAP API
by Marek Grác 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=97… Commit: 975a371ae56179be9f0d806bad7482db8541a4fc Parent: 6ac811961ec028e30cd094bb2a10f6f3e996692d Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 18:24:41 2011 +0200 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Mon Apr 4 18:24:41 2011 +0200 fence_vmware_soap: New fence agent for VMWare using SOAP API This fence agent depends on package python-suds (https://fedorahosted.org/suds/) which provides interface to SOAP Resolves: rhbz#634567 --- fence/agents/vmware_soap/Makefile | 44 ++++++ fence/agents/vmware_soap/fence_vmware_soap.py | 175 +++++++++++++++++++++++++ 2 files changed, 219 insertions(+), 0 deletions(-) diff --git a/fence/agents/vmware_soap/Makefile b/fence/agents/vmware_soap/Makefile new file mode 100644 index 0000000..b657074 --- /dev/null +++ b/fence/agents/vmware_soap/Makefile @@ -0,0 +1,44 @@ +############################################################################### +############################################################################### +## +## Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. +## Copyright (C) 2004 Red Hat, Inc. All rights reserved. +## +## This copyrighted material is made available to anyone wishing to use, +## modify, copy, or redistribute it subject to the terms and conditions +## of the GNU General Public License v.2. +## +############################################################################### +############################################################################### + +SOURCE= fence_vmware_soap.py +TARGET= fence_vmware_soap +MANTARGET= $(TARGET).8 + +top_srcdir=../.. +include ${top_srcdir}/make/defines.mk + +all: $(TARGET) $(MANTARGET) + +fence_vmware_soap: fence_vmware_soap.py + : > $(TARGET) + awk "{print}(\$$1 ~ /#BEGIN_VERSION_GENERATION/){exit 0}" $(SOURCE) >> $(TARGET) + echo "FENCE_RELEASE_NAME=\"${RELEASE}\";" >> $(TARGET) + ${top_srcdir}/scripts/define2var ${top_srcdir}/config/copyright.cf sh REDHAT_COPYRIGHT >> $(TARGET) + echo "BUILD_DATE=\"(built `date`)\";" >> $(TARGET) + awk -v p=0 "(\$$1 ~ /#END_VERSION_GENERATION/){p = 1} {if(p==1)print}" $(SOURCE) >> $(TARGET) + chmod +x $(TARGET) + +install: all + if [ ! -d ${sbindir} ]; then \ + install -d ${sbindir}; \ + fi + install -m755 ${TARGET} ${sbindir} + + if [ ! -d ${mandir}/man8 ]; then \ + install -d ${mandir}/man8; \ + fi + install -m644 ${MANTARGET} ${mandir}/man8 + +clean: + rm -f $(TARGET) $(MANTARGET) diff --git a/fence/agents/vmware_soap/fence_vmware_soap.py b/fence/agents/vmware_soap/fence_vmware_soap.py new file mode 100644 index 0000000..2718560 --- /dev/null +++ b/fence/agents/vmware_soap/fence_vmware_soap.py @@ -0,0 +1,175 @@ +#!/usr/bin/python + +import sys, re, pexpect, exceptions +sys.path.append("@FENCEAGENTSLIBDIR@") + +from suds.client import Client +from suds import WebFault +from suds.sudsobject import Property +from fencing import * + +#BEGIN_VERSION_GENERATION +RELEASE_VERSION="New VMWare Agent - test release on steroids" +REDHAT_COPYRIGHT="" +BUILD_DATE="April, 2011" +#END_VERSION_GENERATION + +def soap_login(options): + if options.has_key("-z"): + url = "https://" + else: + url = "http://" + + url += options["-a"] + ":" + str(options["-u"]) + "/sdk" + conn = Client(url + "/vimService.wsdl") + conn.set_options(location = url) + + mo_ServiceInstance = Property('ServiceInstance') + mo_ServiceInstance._type = 'ServiceInstance' + ServiceContent = conn.service.RetrieveServiceContent(mo_ServiceInstance) + mo_SessionManager = Property(ServiceContent.sessionManager.value) + mo_SessionManager._type = 'SessionManager' + + try: + SessionManager = conn.service.Login(mo_SessionManager, options["-l"], options["-p"]) + except Exception, ex: + fail(EC_LOGIN_DENIED) + + options["ServiceContent"] = ServiceContent + options["mo_SessionManager"] = mo_SessionManager + return conn + +def get_power_status(conn, options): + mo_ViewManager = Property(options["ServiceContent"].viewManager.value) + mo_ViewManager._type = "ViewManager" + + mo_RootFolder = Property(options["ServiceContent"].rootFolder.value) + mo_RootFolder._type = "Folder" + + mo_PropertyCollector = Property(options["ServiceContent"].propertyCollector.value) + mo_PropertyCollector._type = 'PropertyCollector' + + ContainerView = conn.service.CreateContainerView(mo_ViewManager, recursive = 1, container = mo_RootFolder, type = ['VirtualMachine']) + mo_ContainerView = Property(ContainerView.value) + mo_ContainerView._type = "ContainerView" + + FolderTraversalSpec = conn.factory.create('ns0:TraversalSpec') + FolderTraversalSpec.name = "traverseEntities" + FolderTraversalSpec.path = "view" + FolderTraversalSpec.skip = False + FolderTraversalSpec.type = "ContainerView" + + objSpec = conn.factory.create('ns0:ObjectSpec') + objSpec.obj = mo_ContainerView + objSpec.selectSet = [ FolderTraversalSpec ] + objSpec.skip = True + + propSpec = conn.factory.create('ns0:PropertySpec') + propSpec.all = False + propSpec.pathSet = ["name", "summary.runtime.powerState", "config.uuid", "summary", "config", "capability", "network"] + propSpec.type = "VirtualMachine" + + propFilterSpec = conn.factory.create('ns0:PropertyFilterSpec') + propFilterSpec.propSet = [ propSpec ] + propFilterSpec.objectSet = [ objSpec ] + + try: + raw_machines = conn.service.RetrievePropertiesEx(mo_PropertyCollector, propFilterSpec) + except Exception, ex: + fail(EC_STATUS) + + machines = { } + uuid = { } + mappingToUUID = { } + + for m in raw_machines.objects: + info = {} + for i in m.propSet: + info[i.name] = i.val + machines[info["name"]] = (info["config.uuid"], info["summary.runtime.powerState"]) + uuid[info["config.uuid"]] = info["summary.runtime.powerState"] + mappingToUUID[m.obj.value] = info["config.uuid"] + + if ["list", "monitor"].count(options["-o"]) == 1: + return machines + else: + if options.has_key("-U") == False: + ## Transform InventoryPath to UUID + mo_SearchIndex = Property(options["ServiceContent"].searchIndex.value) + mo_SearchIndex._type = "SearchIndex" + + vm = conn.service.FindByInventoryPath(mo_SearchIndex, options["-n"]) + + try: + options["-U"] = mappingToUUID[vm.value] + except KeyError, ex: + fail(EC_STATUS) + except AttributeError, ex: + fail(EC_STATUS) + + try: + if uuid[options["-U"]] == "poweredOn": + return "on" + else: + return "off" + return status + except KeyError, ex: + fail(EC_STATUS) + +def set_power_status(conn, options): + mo_SearchIndex = Property(options["ServiceContent"].searchIndex.value) + mo_SearchIndex._type = "SearchIndex" + vm = conn.service.FindByUuid(mo_SearchIndex, vmSearch = 1, uuid = options["-U"]) + + mo_machine = Property(vm.value) + mo_machine._type = "VirtualMachine" + + if options["-o"] == "on": + conn.service.PowerOnVM_Task(mo_machine) + else: + conn.service.PowerOffVM_Task(mo_machine) + +def main(): + device_opt = [ "help", "version", "agent", "quiet", "verbose", "debug", + "action", "ipaddr", "login", "passwd", "passwd_script", + "ssl", "port", "uuid", "separator", "ipport", + "power_timeout", "shell_timeout", "login_timeout", "power_wait" ] + + atexit.register(atexit_handler) + + options = check_input(device_opt, process_input(device_opt)) + + ## + ## Fence agent specific defaults + ##### + docs = { } + docs["shortdesc"] = "Fence agent for VMWare over SOAP API" + docs["longdesc"] = "fence_vmware_soap is an I/O Fencing agent \ +which can be used with the virtual machines managed by VMWare products \ +that have SOAP API v4.1+. \ +\n.P\n\ +Name of virtual machine (-n / port) has to be used in inventory path \ +format (e.g. /datacenter/vm/Discovered virtual machine/myMachine). Alternatively \ +you can use UUID (-U / uuid) to access virtual machine." + docs["vendorurl"] = "http://www.vmware.com" + show_docs(options, docs) + + ## + ## Operate the fencing device + #### + conn = soap_login(options) + + result = fence_action(conn, options, set_power_status, get_power_status, get_power_status) + + ## + ## Logout from system + ##### + try: + conn.service.Logout(options["mo_SessionManager"]) + except Exception, ex: + pass + + sys.exit(result) + +if __name__ == "__main__": + main()

1 0

cluster: RHEL57 - library: Add support for UUID (-U / --uuid / uuid)
by Marek Grác 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=6a… Commit: 6ac811961ec028e30cd094bb2a10f6f3e996692d Parent: 48f5feec3bbd0375dff3024ce4deca7090608325 Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 18:19:40 2011 +0200 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Mon Apr 4 18:19:40 2011 +0200 library: Add support for UUID (-U / --uuid / uuid) Fence agents can use --uuid as a replacement for --port. This is very useful in a case of virtual machines that have both UUID and name. Patch proposed by Matt Clark --- fence/agents/lib/fencing.py.py | 13 ++++++++++--- 1 files changed, 10 insertions(+), 3 deletions(-) diff --git a/fence/agents/lib/fencing.py.py b/fence/agents/lib/fencing.py.py index 933fd75..61c23af 100644 --- a/fence/agents/lib/fencing.py.py +++ b/fence/agents/lib/fencing.py.py @@ -390,7 +390,14 @@ all_opt = { "default" : "1", "required" : "0", "shortdesc" : "Count of attempts to retry power on", - "order" : 200 } + "order" : 201 }, + "uuid" : { + "getopt" : "U:", + "longopt" : "uuid", + "help" : "-U, --uuid UUID of the VM to fence.", + "required" : "0", + "shortdesc" : "The UUID of the virtual machine to fence.", + "order" : 1 } } common_opt = [ "retry_on", "delay" ] @@ -691,7 +698,7 @@ def check_input(device_opt, opt): if 0 == os.path.isfile(options["-k"]): fail_usage("Failed: Identity file " + options["-k"] + " does not exist") - if (0 == ["list", "monitor"].count(options["-o"].lower())) and (0 == options.has_key("-n")) and (device_opt.count("port")): + if (0 == ["list", "monitor"].count(options["-o"].lower())) and (0 == options.has_key("-n") and 0 == options.has_key("-U")) and (device_opt.count("port")): fail_usage("Failed: You have to enter plug number") if options.has_key("-S"): @@ -757,7 +764,7 @@ def fence_action(tn, options, set_power_fn, get_power_fn, get_outlet_list = None ## Process options that manipulate fencing device ##### - if (options["-o"] == "list") and (0 == options["device_opt"].count("port")) and (0 == options["device_opt"].count("partition")): + if (options["-o"] == "list") and (0 == options["device_opt"].count("port")) and (0 == options["device_opt"].count("partition") and 0 == options["device_opt"].count("uuid")): print "N/A" return elif (options["-o"] == "list" and get_outlet_list == None):

1 0

cluster: RHEL57 - fence_ipmilan: returns incorrect status on monitor op if chassis is powered off
by Marek Grác 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=48… Commit: 48f5feec3bbd0375dff3024ce4deca7090608325 Parent: 3f19f28cc76c564f845fddd54a231f35410015fa Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 15:11:59 2011 +0200 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Mon Apr 4 18:15:32 2011 +0200 fence_ipmilan: returns incorrect status on monitor op if chassis is powered off There is a bug in fence_ipmilan agent: it exits with result code 2 on "monitor"/"status" command if chassis is powered off. Instead it should return 0 because "monitor" command should return fence device status, not chassis status. Reported and patch proposed by Vladislav Bogdanov Resolves: rhbz#693427 --- fence/agents/ipmilan/ipmilan.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/fence/agents/ipmilan/ipmilan.c b/fence/agents/ipmilan/ipmilan.c index 4371143..7031982 100644 --- a/fence/agents/ipmilan/ipmilan.c +++ b/fence/agents/ipmilan/ipmilan.c @@ -1445,6 +1445,10 @@ metaout: else printf("Failed\n"); } + + if (!strcasecmp(op, "monitor")) + translated_ret = ret; + return translated_ret; } #endif /* fence */

1 0

cluster: RHEL57 - fence_cisco_ucs, fence_rhevm: Problems with SSL support
by Marek Grác 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=3f… Commit: 3f19f28cc76c564f845fddd54a231f35410015fa Parent: 56ac82f2ed4babb8940305411591234698ca212c Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Apr 4 14:59:47 2011 +0200 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Mon Apr 4 18:13:56 2011 +0200 fence_cisco_ucs, fence_rhevm: Problems with SSL support Both of these agents are using pycurl. This library has different default setting across various distributions, so it is better to set SSL options directly. Resolves: rhbz#691392 --- fence/agents/cisco_ucs/fence_cisco_ucs.py | 1 + fence/agents/rhevm/fence_rhevm.py | 1 + 2 files changed, 2 insertions(+), 0 deletions(-) diff --git a/fence/agents/cisco_ucs/fence_cisco_ucs.py b/fence/agents/cisco_ucs/fence_cisco_ucs.py index 8505c31..beecd3a 100755 --- a/fence/agents/cisco_ucs/fence_cisco_ucs.py +++ b/fence/agents/cisco_ucs/fence_cisco_ucs.py @@ -88,6 +88,7 @@ def send_command(opt, command): c.setopt(pycurl.WRITEFUNCTION, b.write) c.setopt(pycurl.TIMEOUT, int(opt["-Y"])) c.setopt(pycurl.SSL_VERIFYPEER, 0) + c.setopt(pycurl.SSL_VERIFYHOST, 0) c.perform() result = b.getvalue() diff --git a/fence/agents/rhevm/fence_rhevm.py b/fence/agents/rhevm/fence_rhevm.py index 323ccf8..1cb8d00 100755 --- a/fence/agents/rhevm/fence_rhevm.py +++ b/fence/agents/rhevm/fence_rhevm.py @@ -99,6 +99,7 @@ def send_command(opt, command, method = "GET"): c.setopt(pycurl.USERPWD, opt["-l"] + ":" + opt["-p"]) c.setopt(pycurl.TIMEOUT, int(opt["-Y"])) c.setopt(pycurl.SSL_VERIFYPEER, 0) + c.setopt(pycurl.SSL_VERIFYHOST, 0) if (method == "POST"): c.setopt(pycurl.POSTFIELDS, "<action />")

1 0

fence-agents: the annotated tag v3.1.3 has been created
by Fabio M. Di Nitto 04 Apr '11

04 Apr '11

Gitweb: http://git.fedorahosted.org/git/fence-agents.git?p=fence-agents.git;a=commi… Commit: e06cfed120211198c0a1d032a1ece715bc8fe46b Parent: 0000000000000000000000000000000000000000 Author: Fabio M. Di Nitto <fdinitto(a)redhat.com> AuthorDate: 2011-04-04 15:24 +0000 Committer: Fabio M. Di Nitto <fdinitto(a)redhat.com> CommitterDate: 2011-04-04 15:24 +0000 annotated tag: v3.1.3 has been created at e06cfed120211198c0a1d032a1ece715bc8fe46b (tag) tagging bf66212512ad7f97b38fcb49fc256ac76583afbd (commit) replaces v3.1.2 v3.1.3 release Arnaud Quette (1): eaton_snmp: fix list offset Brandon Perkins (1): Fence_rhevm needs to change "RUNNING" status to "UP" status as the REST-API has changed. Jim Ramsay (1): Allow fence_scsi to use any valid hexadecimal key Lon Hohberger (1): fence-agents: Accept other values for "true" Marek 'marx' Grac (10): fence_rhevm: Update URL to RHEV-M REST API fence_cisco_ucs: Support for sub-organization fence_cisco_ucs: Fix for support for sub-organization library: Add support for UUID (-U / --uuid / uuid) fence_vmware_soap: New fence agent for VMWare using SOAP API fence_cisco_ucs, fence_rhevm: Problems with SSL support fence_ipmilan: returns incorrect status on monitor op if chassis is powered off fence_ipmilan: Correct return code for diag operation library: Add support for URL session instead of hostname bugfix uuid Ryan O'Hara (2): fence_scsi: fix typo when opening logfile fence_scsi: grep for keys should be case insensitive

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

cluster-commits April 2011