cluster-commits July 2010

cluster-commits@lists.fedorahosted.org

9 participants
175 discussions

fence-agents: master - fence_drac5: make "port" a synonym of "module_name" for drac5
by Marek Grác 30 Jul '10

30 Jul '10

Gitweb: http://git.fedorahosted.org/git/fence-agents.git?p=fence-agents.git;a=commi… Commit: 0e84b2e3f6ef1be68cbdb2414eb257f879f43af7 Parent: d1128ab1406df77566bc91db28d1f97f6ba101cb Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Fri Jul 30 17:58:01 2010 +0200 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Fri Jul 30 17:58:01 2010 +0200 fence_drac5: make "port" a synonym of "module_name" for drac5 Resolves: rhbz#619776 --- fence/agents/lib/fencing.py.py | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-) diff --git a/fence/agents/lib/fencing.py.py b/fence/agents/lib/fencing.py.py index 7d967de..4c5a0db 100644 --- a/fence/agents/lib/fencing.py.py +++ b/fence/agents/lib/fencing.py.py @@ -521,6 +521,11 @@ def process_input(avail_opt): else: longopt_list.append(all_opt[k]["longopt"]) + ## Compatibility layer + if avail_opt.count("module_name") == 1: + getopt_string += "n:" + longopt_list.append("plug=") + ## ## Read options from command line or standard input ##### @@ -547,6 +552,8 @@ def process_input(avail_opt): z = dict(opt) if z.has_key("-T") == 1: z["-o"] = "status" + if z.has_key("-n") == 1: + z["-m"] = z["-n"] opt = z ## @@ -576,6 +583,8 @@ def process_input(avail_opt): name = "module_name" elif name == "action" and 1 == avail_opt.count("io_fencing"): name = "io_fencing" + elif name == "port" and 1 == avail_opt.count("drac_version"): + name = "module_name" ## ######

1 0

cluster: the annotated tag cluster-3.0.14 has been created
by Fabio M. Di Nitto 30 Jul '10

30 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=55… Commit: 55dd07589f0ace485c31905a6b0fc8538a00c2eb Parent: 0000000000000000000000000000000000000000 Author: Fabio M. Di Nitto <fdinitto(a)redhat.com> AuthorDate: 2010-07-30 11:46 +0000 Committer: Fabio M. Di Nitto <fdinitto(a)redhat.com> CommitterDate: 2010-07-30 11:46 +0000 annotated tag: cluster-3.0.14 has been created at 55dd07589f0ace485c31905a6b0fc8538a00c2eb (tag) tagging db1a93f32b36da95a19716ed1b6832b04f4c68dd (commit) replaces cluster-3.0.13 cluster-3.0.14 release Abhijith Das (1): gfs2 manual pages: gfs2_convert manpage and documentation updates Bob Peterson (4): gfs2_edit restoremeta should not return 0 on failure fsck.gfs2: unaligned access on ia64 GFS2: libgfs2 bitfit algorithm using wrong shift point Make gfs2_edit show bit-to-block translation when viewing bitmaps Carlos Eduardo Maiolino (1): resource-agents: Remove nfs service temp directories Christine Caulfield (2): cman: Recalculate expected_votes on a config reload. cman: Mention cman_tool version -S in man page David Teigland (2): dlm_controld: fix plock checkpoint signatures dlm_controld: fix plock owner in checkpoints Fabio M. Di Nitto (17): resource agents: Remove bashisms from resource scripts cman init: more LSB compliance rgmanger init: more LSB complaint bits cman: add default config snippet for cman init script gfs2 init: make the init script LSB compliant cman config: copy all logging objects to the top level tree fence: rename ibmblade to bladecenter_snmp fence agents: add compatibility symlink for ibmblade / bladecenter_snmp cman: allow init script to pass options to fenced config: fix several issues with reload operation cman-preconfig: better handle of logging reload operation cman-preconfig: better handle of logging reload operation (part 2) config: free new config if we cannot find the config_version config: more cman_tool config reload cleanup cman: simply message broadcasting handling config: better error report when autodetecting config version errors cman: fix consensus calculation Federico Simoncelli (1): config: Add missing cman_label Guido Günther (1): fence_rsb: Raise exceptions not strings Jan Friesse (1): cman: check for new config only once per second Lon Hohberger (37): config: Add missing resource docs to cluster.rng config: Clean up recursion and documentation config: Add documentation for interface tag Revert "config: Clean up recursion and documentation" Revert "config: Add missing resource docs to cluster.rng" resource-agents: Add missing resource docs config: Add missing resource docs to cluster.rng config: Clean up recursion and documentation resource-agents: Clean up recursion and documentation rgmanager: Pass timeouts to resource agents resource-agents: Make vm.sh use stop/start timeouts rgmanager: Use sysrq-b to reboot Revert "resource-agents: Make vm.sh use stop/start timeouts" resource-agents: Make vm.sh use stop/start timeouts resource-agents: fix incorrect link resolution in fs-lib rgmanager: Make clustat -f not query CCS/objdb resource-agents: Add NFSv4 support resource-agents: Add NFSv4 agent to installation config: Update cluster schema config: Update cluster LDIF schema doc: Add auto-generated cluster.conf reference doc: Install cluster.conf reference rgmanager: Make clulog filter correctly resource-agents: Add resource type to logging rgmanager: Man page improvements resource-agents: Allow other values for "yes" config: Allow multiple logging_daemon tags config: Add doc for cman_label attribute config: Update LDIF schema cman: Recalculate quorum on config change config: Add tomcat-6 resource agent to schema config: Add tomcat-6 to ldif schema doc: Add tomcat-6 to cluster_conf.html cman: Recalculate quorum on quorum device vote changes cman: Deprecate specifying config version to cman_tool config: Regenerate fencing definitions for XML schema config: Exclude fence_scsi from generated section Marek 'marx' Grac (8): fence_wti: support non-default TCP ports fence_apc: fence_apc fails for some port numbers resource-agents: Use SIGQUIT if SIGTERM was not fast enough resource-agents: new agent for tomcat 6 resource-agents: change build system to include tomcat6 RA resource-agents: RA for psql does not work correctly with netmask fence_ilo: will throw exception if user does not have power priviledges fencing: Not all parameters appear in metadata Masahiro Matsuya (1): resource-agents: Fix migration mapping behavior w/ virsh Ryan O'Hara (1): Fix open flags so that logfile won't be truncated each time we open it.

1 0

cluster: RHEL56 - qdisk: mkqdisk argument positioning
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=c6… Commit: c6873545f430825a0f211dd595a2248f24ab88c1 Parent: eff6913e7cea3be112e6635c168d2215c9735d2b Author: Tatsuo Kawasaki <kawasaki(a)redhat.com> AuthorDate: Fri Feb 12 14:42:52 2010 -0500 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 16:01:46 2010 -0400 qdisk: mkqdisk argument positioning mkqdisk debug option (-d) doesn't work after -L option. This works: mkqdisk -d -L This should also work: mkqdisk -L -d This patch fixes the behavior. Resolves: rhbz#555246 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- cman/qdisk/mkqdisk.c | 21 ++++++++++++++------- 1 files changed, 14 insertions(+), 7 deletions(-) diff --git a/cman/qdisk/mkqdisk.c b/cman/qdisk/mkqdisk.c index 46d6b63..edccf1c 100644 --- a/cman/qdisk/mkqdisk.c +++ b/cman/qdisk/mkqdisk.c @@ -37,7 +37,7 @@ main(int argc, char **argv) { char device[128]; char *newdev = NULL, *newlabel = NULL; - int rv, verbose_level = 1; + int rv, flg = 0, verbose_level = 1; printf("mkqdisk v0.6.0\n"); @@ -48,13 +48,12 @@ main(int argc, char **argv) break; case 'L': /* List */ - close(2); - return find_partitions(NULL, NULL, 0, verbose_level); - break; + flg = rv; + break; case 'f': - close(2); - return find_partitions( optarg, device, - sizeof(device), verbose_level); + flg = rv; + newlabel = optarg; + break; case 'c': newdev = optarg; break; @@ -70,6 +69,14 @@ main(int argc, char **argv) } } + /* list */ + if (flg == 'L') { + return find_partitions(NULL, NULL, 0, verbose_level); + } else if (flg == 'f') { + return find_partitions( newlabel, device, + sizeof(device), verbose_level); + } + if (!newdev && !newlabel) { printf("usage: mkqdisk -L | -f <label> | -c " "<device> -l <label>\n");

1 0

cluster: RHEL56 - ccsd: Add error when bind fails
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=ef… Commit: eff6913e7cea3be112e6635c168d2215c9735d2b Parent: fa36367db8a3650a1cc761b7570ef2b4c0b3b88b Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Thu Jul 29 15:53:59 2010 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 15:53:59 2010 -0400 ccsd: Add error when bind fails Resolves: rhbz#573996 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- ccs/daemon/cluster_mgr.c | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-) diff --git a/ccs/daemon/cluster_mgr.c b/ccs/daemon/cluster_mgr.c index 7cfa531..3fb07fe 100644 --- a/ccs/daemon/cluster_mgr.c +++ b/ccs/daemon/cluster_mgr.c @@ -364,6 +364,7 @@ static int handle_cluster_event(cman_handle_t handle) static void cluster_communicator(void) { + char addrbuf[64], *res; int ccsd_fd = -1; int cman_fd = -1; int warn_user = 0; @@ -398,7 +399,12 @@ static void cluster_communicator(void) fcntl(ccsd_fd, F_SETFD, flags); if (bind(ccsd_fd, (struct sockaddr *)&sin, sizeof(sin)) < 0) { - log_err("Unable to bind to socket.\n"); + res = inet_ntop(AF_INET,&sin.sin_addr, addrbuf, sizeof(addrbuf)); + if (!res) { + log_sys_err("Unable to bind socket addr: %s, port: %d, proto: %d", addrbuf, sin.sin_port, sin.sin_family); + } else { + log_sys_err("Unable to bind to ipv4 socket, but inet_ntop returned NULL pointer"); + } close(ccsd_fd); exit(EXIT_FAILURE); }

1 0

cluster: RHEL56 - fence-agents: Update fence_ack_manual man page
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=fa… Commit: fa36367db8a3650a1cc761b7570ef2b4c0b3b88b Parent: b2277bbc5295f52944a6901695cee71c7f0ee018 Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Thu Jul 29 15:47:26 2010 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 15:47:26 2010 -0400 fence-agents: Update fence_ack_manual man page Resolves: rhbz#578604 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- fence/man/fence_ack_manual.8 | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/fence/man/fence_ack_manual.8 b/fence/man/fence_ack_manual.8 index fa9c4ed..29f947f 100644 --- a/fence/man/fence_ack_manual.8 +++ b/fence/man/fence_ack_manual.8 @@ -23,6 +23,10 @@ then run fence_ack_manual. Running fence_ack_manual allows the cluster to continue with recovery of the fenced machine. The victim may be disconnected from storage rather than resetting it. +fence_ack_manual may also be used to allow a cluster to resume operation +after fencing has failed for a host. This is not related to fence_manual(8), +requires no configuration, and may be used whenever fencing has failed. + .SH OPTIONS .TP \fB-h\fP @@ -37,6 +41,10 @@ Name of node that has been reset or disconnected from storage. \fB-s\fP \fIIPaddress\fP IP address of the machine which has been reset or disconnected from storage. (Deprecated; use -n instead.) .TP +\fB-e\fP +Emergency fencing override. This may be used when fencing a given host +is failing in order to restore the cluster to operation. +.TP \fB-V\fP Print out a version message, then exit. .SH SEE ALSO

1 0

cluster: RHEL56 - resource-agents: SAPDatabase: remove $TEMPFILE
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=b2… Commit: b2277bbc5295f52944a6901695cee71c7f0ee018 Parent: 94085eace39e248040cf7069c7294178c6f944ce Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Mon Feb 15 16:14:50 2010 -0500 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 14:16:00 2010 -0400 resource-agents: SAPDatabase: remove $TEMPFILE Resolves: rhbz#565363 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- rgmanager/src/resources/SAPDatabase | 52 +++++++++++++++++++++++------------ 1 files changed, 34 insertions(+), 18 deletions(-) diff --git a/rgmanager/src/resources/SAPDatabase b/rgmanager/src/resources/SAPDatabase index d6972fd..8ac3005 100644 --- a/rgmanager/src/resources/SAPDatabase +++ b/rgmanager/src/resources/SAPDatabase @@ -184,6 +184,19 @@ trap_handler() { } +do_exit() { + # If we've got a tempfile variable and the tempfile exists... + # ... if the return code is 0 *or* the temp file is empty + # remove it. + if [ -n "$TEMPFILE" ] && [ -e "$TEMPFILE" ]; then + if [ $1 -eq 0 ] || [ "$(stat -c %s $TEMPFILE)" = "0" ]; then + rm -f $TEMPFILE + fi + fi + exit $1 +} + + # # listener_start: Start the given listener # @@ -822,10 +835,6 @@ then exit $OCF_ERR_ARGS fi -# Set a tempfile and make sure to clean it up again -TEMPFILE="$(mktemp /tmp/SAPDatabase.tmp.XXXXXX)" -trap trap_handler INT TERM - # These operations don't require OCF instance parameters to be set case "$1" in meta-data) meta_data @@ -840,25 +849,32 @@ case "$1" in *);; esac +# Set a tempfile and make sure to clean it up again +TEMPFILE="$(mktemp /tmp/SAPDatabase.tmp.XXXXXX)" +trap trap_handler INT TERM + +# Everything after here must call do_exit to remove temp file + + US=`id -u -n` US=`echo $US` if [ $US != root ] then ocf_log err "$0 must be run as root" - exit $OCF_ERR_PERM + do_exit $OCF_ERR_PERM fi # mandatory parameter check if [ -z "$OCF_RESKEY_SID" ]; then ocf_log err "Please set OCF_RESKEY_SID to the SAP system id!" - exit $OCF_ERR_ARGS + do_exit $OCF_ERR_ARGS fi SID=`echo "$OCF_RESKEY_SID"` if [ -z "$OCF_RESKEY_DBTYPE" ]; then ocf_log err "Please set OCF_RESKEY_DBTYPE to the database vendor specific tag (ORA,ADA,DB6)!" - exit $OCF_ERR_ARGS + do_exit $OCF_ERR_ARGS fi DBTYPE=`echo "$OCF_RESKEY_DBTYPE" | tr "[a-z]" "[A-Z]"` @@ -877,7 +893,7 @@ else ;; 0|false|FALSE|no|NO) DBJ2EE_ONLY=0;; *) ocf_log err "Parsing parameter DBJ2EE_ONLY: '$DBJ2EE_ONLY' is not a boolean value!" - exit $OCF_ERR_ARGS ;; + do_exit $OCF_ERR_ARGS ;; esac fi @@ -897,7 +913,7 @@ else 1|true|TRUE|yes|YES) OCF_RESKEY_STRICT_MONITORING=1;; 0|false|FALSE|no|NO) OCF_RESKEY_STRICT_MONITORING=0;; *) ocf_log err "Parsing parameter STRICT_MONITORING: '$OCF_RESKEY_STRICT_MONITORING' is not a boolean value!" - exit $OCF_ERR_ARGS ;; + do_exit $OCF_ERR_ARGS ;; esac fi @@ -922,7 +938,7 @@ done if [ -z "$DIR_EXECUTABLE" ] then ocf_log warn "Cannot find $EXESTARTDB,$EXESTOPDB and $EXEDBCONNECT executable, please set DIR_EXECUTABLE parameter!" - exit $OCF_NOT_RUNNING + do_exit $OCF_NOT_RUNNING fi if [ $DBJ2EE_ONLY -eq 1 ] @@ -951,7 +967,7 @@ then PATH=$JAVA_HOME/bin:$PATH else ocf_log err "Cannot find JAVA_HOME directory, please set JAVA_HOME parameter!" - exit $OCF_NOT_RUNNING + do_exit $OCF_NOT_RUNNING fi fi @@ -985,25 +1001,25 @@ sidadm="`echo $SID | tr [:upper:] [:lower:]`adm" case "$1" in start) sapdatabase_start - exit $?;; + do_exit $?;; stop) sapdatabase_stop - exit $?;; + do_exit $?;; monitor) sapdatabase_monitor $OCF_RESKEY_STRICT_MONITORING - exit $?;; + do_exit $?;; status) sapdatabase_status - exit $?;; + do_exit $?;; recover) sapdatabase_recover - exit $?;; + do_exit $?;; validate-all) sapdatabase_validate - exit $?;; + do_exit $?;; *) sapdatabase_methods - exit $OCF_ERR_UNIMPLEMENTED;; + do_exit $OCF_ERR_UNIMPLEMENTED;; esac

1 0

cluster: RHEL56 - resource-agents: Fix migration mapping behavior w/ virsh
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=94… Commit: 94085eace39e248040cf7069c7294178c6f944ce Parent: 8dff6d4626831bf941a32ee75e9b802fc51a0e8f Author: Masahiro Matsuya <mmatsuya(a)redhat.com> AuthorDate: Fri Jul 9 14:40:51 2010 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 14:13:46 2010 -0400 resource-agents: Fix migration mapping behavior w/ virsh Consider a two node cluster. The hostname of the nodes are 'sk010001' and 'sk010002'. Each nodes has two bonded network interfaces for public and private communications. The hostname matches the hostname of the IP address on public network. Node1: sk010001 bond0 (for public network) : 172.22.51.1 sk010001 bond2 (for private network): 172.22.48.131 sk010001-hb Node2: sk010002 bond0 (for public network) : 172.22.51.2 sk010002 bond2 (for private network): 172.22.48.132 sk010002-hb In cluster.conf, a migration mapping is used to specify that the private interfaces should be used for migration traffic. Unfortunately, when doing a live migration, while the traffic should use the -hb interfaces, bond0 is used. This is because the vm.sh agent uses the following command for live migration from sk010001 to sk010002: virsh migrate --live su21k003 \ qemu+ssh://sk010002-hb/system This is not enough to ensure the guest goes over the private interface. The --migrateuri option of 'virsh migrate' is needed for it. The following command should be executed instead: virsh migrate --live su21k003 \ qemu+ssh://sk010002-hb/system tcp:sk010002-hb Resolves: rhbz#596016 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- rgmanager/src/resources/vm.sh | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/rgmanager/src/resources/vm.sh b/rgmanager/src/resources/vm.sh index c8be516..61ced72 100755 --- a/rgmanager/src/resources/vm.sh +++ b/rgmanager/src/resources/vm.sh @@ -787,6 +787,7 @@ validate_all() # Virsh makes it easier to do this. Really. if [ "$OCF_RESKEY_hypervisor" = "qemu" ]; then export OCF_RESKEY_migration_uri="qemu+ssh://%s/system" + export migrateuriopt="tcp:%s" fi # I just need to believe in it more. @@ -822,7 +823,7 @@ virsh_migrate() err=$($cmd 2>&1 | head -1; exit ${PIPESTATUS[0]}) rv=$? elif [ "$OCF_RESKEY_hypervisor" = "qemu" ]; then - cmd="virsh migrate $migrate_opt $OCF_RESKEY_name $(printf $OCF_RESKEY_migration_uri $target)" + cmd="virsh migrate $migrate_opt $OCF_RESKEY_name $(printf $OCF_RESKEY_migration_uri $target) $(printf $migrateuriopt $target)" ocf_log debug "$cmd" err=$($cmd 2>&1 | head -1; exit ${PIPESTATUS[0]})

1 0

cluster: RHEL56 - rgmanager: Kill processes correctly w/ force_unmount
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=8d… Commit: 8dff6d4626831bf941a32ee75e9b802fc51a0e8f Parent: fbc82625c84ca9dfd99f7d8e3e051c53a63bf523 Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Wed Apr 14 17:29:54 2010 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 14:06:48 2010 -0400 rgmanager: Kill processes correctly w/ force_unmount The killMountProcesses function was written about 10 years ago. It was designed to work with lsof or fuser, and to log messages for each process killed. This is not a bad idea. The problem is that parsing the output of either is and error-prone, particularly when mountpoints are similar to other directories on the system. A far less error-prone method to cleaning up a mount point is to use 'fuser -kvm' on it. Not only is this less error-prone, it's a good bit faster at doing its job than iterating through output in a shell script. This patch makes force_unmount very reliable at killing the correct processes, but we lose the logging functionality. It is a fair trade-off because there have been several bugs in the killMountProcesses function over the years which have caused several problems. Resolves: bz573705 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- rgmanager/src/resources/clusterfs.sh | 113 ++---------------------------- rgmanager/src/resources/fs.sh | 124 ++------------------------------- rgmanager/src/resources/netfs.sh | 128 ++++------------------------------ 3 files changed, 25 insertions(+), 340 deletions(-) diff --git a/rgmanager/src/resources/clusterfs.sh b/rgmanager/src/resources/clusterfs.sh index 250978e..6a837b1 100755 --- a/rgmanager/src/resources/clusterfs.sh +++ b/rgmanager/src/resources/clusterfs.sh @@ -504,113 +504,6 @@ isAlive() # -# killMountProcesses device mount_point -# -# Using lsof or fuser try to unmount the mount by killing of the processes -# that might be keeping it busy. -# -killMountProcesses() -{ - typeset -i ret=$SUCCESS - typeset have_lsof="" - typeset have_fuser="" - typeset try - - if [ $# -ne 1 ]; then - ocf_log err \ - "Usage: killMountProcesses mount_point" - return $FAIL - fi - - typeset mp=$1 - - ocf_log notice "Forcefully unmounting $mp" - - # - # Not all distributions have lsof. If not use fuser. If it - # does, try both. - # - file=$(which lsof 2>/dev/null) - if [ -f "$file" ]; then - have_lsof=$YES - fi - - file=$(which fuser 2>/dev/null) - if [ -f "$file" ]; then - have_fuser=$YES - fi - - if [ -z "$have_lsof" -a -z "$have_fuser" ]; then - ocf_log warn \ - "Cannot forcefully unmount $mp; cannot find lsof or fuser commands" - return $FAIL - fi - - for try in 1 2 3; do - if [ -n "$have_lsof" ]; then - # - # Use lsof to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mountpoint $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(lsof -b 2>/dev/null | \ - grep -E "$mp(/.*|)\$" | \ - awk '{print $1,$2,$3}' | \ - sort -u -k 1,3) - elif [ -n "$have_fuser" ]; then - # - # Use fuser to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mount point $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(fuser -vm $mp 2>&1 | \ - grep -v PID | \ - sed 's;^'$mp:';;' | \ - awk '{print $4,$2,$1}' | \ - sort -u -k 1,3) - fi - done - - return $ret -} - -# # startFilesystem # startFilesystem() { @@ -891,7 +784,11 @@ stop: Could not match $OCF_RESKEY_device with a real device" umount_failed=yes if [ "$force_umount" ]; then - killMountProcesses $mp + if [ $try -eq 1 ]; then + fuser -TERM -kvm "$mp" + else + fuser -kvm "$mp" + fi fi if [ $try -ge $max_tries ]; then diff --git a/rgmanager/src/resources/fs.sh b/rgmanager/src/resources/fs.sh index 900dca5..a2148f4 100755 --- a/rgmanager/src/resources/fs.sh +++ b/rgmanager/src/resources/fs.sh @@ -695,115 +695,8 @@ isAlive() # -# killMountProcesses mount_point -# -# Using lsof or fuser try to unmount the mount by killing of the processes -# that might be keeping it busy. -# -killMountProcesses() -{ - typeset -i ret=$SUCCESS - typeset have_lsof="" - typeset have_fuser="" - typeset try - - if [ $# -ne 1 ]; then - ocf_log err \ - "Usage: killMountProcesses mount_point" - return $FAIL - fi - - typeset mp=$1 - - ocf_log notice "Forcefully unmounting $mp" - - # - # Not all distributions have lsof. If not use fuser. If it - # does, try both. - # - file=$(which lsof 2>/dev/null) - if [ -f "$file" ]; then - have_lsof=$YES - fi - - file=$(which fuser 2>/dev/null) - if [ -f "$file" ]; then - have_fuser=$YES - fi - - if [ -z "$have_lsof" -a -z "$have_fuser" ]; then - ocf_log warn \ - "Cannot forcefully unmount $mp; cannot find lsof or fuser commands" - return $FAIL - fi - - for try in 1 2 3; do - if [ -n "$have_lsof" ]; then - # - # Use lsof to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mountpoint $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(lsof -bn 2>/dev/null | \ - grep -E " $mp(/| |$)" | \ - awk '{print $1,$2,$3}' | \ - sort -u -k 1,3) - elif [ -n "$have_fuser" ]; then - # - # Use fuser to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mount point $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(fuser -vm $mp 2>&1 | \ - grep -v PID | \ - sed 's;^'$mp:';;' | \ - awk '{print $4,$2,$1}' | \ - sort -u -k 1,3) - fi - done - - return $ret -} - - -# -# Enable quotas on the mount point if the user requested them +# Decide which quota options are enabled and return a string +# which we can pass to quotaon # enable_fs_quotas() { @@ -1176,8 +1069,9 @@ stop: Could not match $OCF_RESKEY_device with a real device" umount_failed=yes if [ "$force_umount" ]; then - killMountProcesses $mp if [ $try -eq 1 ]; then + fuser -TERM -kvm "$mp" + if [ "$OCF_RESKEY_nfslock" = "yes" ] || \ [ "$OCF_RESKEY_nfslock" = "1" ]; then ocf_log warning \ @@ -1189,15 +1083,11 @@ stop: Could not match $OCF_RESKEY_device with a real device" notify_list_store $mp/.clumanager/statd nfslock_reclaim=1 fi + else + fuser -kvm "$mp" fi fi - if [ $try -ge $max_tries ]; then - done=$YES - else - sleep $sleep_time - let try=try+1 - fi ;; *) return $FAIL @@ -1206,7 +1096,7 @@ stop: Could not match $OCF_RESKEY_device with a real device" if [ $try -ge $max_tries ]; then done=$YES - elif [ "$done" -ne "$YES" ]; then + elif [ "$done" != "$YES" ]; then sleep $sleep_time let try=try+1 fi diff --git a/rgmanager/src/resources/netfs.sh b/rgmanager/src/resources/netfs.sh index bd391b6..2cc6863 100755 --- a/rgmanager/src/resources/netfs.sh +++ b/rgmanager/src/resources/netfs.sh @@ -359,113 +359,6 @@ isMounted () { } # -# killMountProcesses mount_point -# -# Using lsof or fuser try to unmount the mount by killing of the processes -# that might be keeping it busy. -# -killMountProcesses() -{ - typeset -i ret=$SUCCESS - typeset have_lsof="" - typeset have_fuser="" - typeset try - - if [ $# -ne 1 ]; then - ocf_log err \ - "Usage: killMountProcesses mount_point" - return $FAIL - fi - - typeset mp=$1 - - ocf_log notice "Forcefully unmounting $mp" - - # - # Not all distributions have lsof. If not use fuser. If it - # does, try both. - # - file=$(which lsof 2>/dev/null) - if [ -f "$file" ]; then - have_lsof=$YES - fi - - file=$(which fuser 2>/dev/null) - if [ -f "$file" ]; then - have_fuser=$YES - fi - - if [ -z "$have_lsof" -a -z "$have_fuser" ]; then - ocf_log warn \ - "Cannot forcefully unmount $mp; cannot find lsof or fuser commands" - return $FAIL - fi - - for try in 1 2 3; do - if [ -n "$have_lsof" ]; then - # - # Use lsof to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mountpoint $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(lsof -w -bn 2>/dev/null | \ - grep -w -E "$mp(/.*|)\$" | \ - awk '{print $1,$2,$3}' | \ - sort -u -k 1,3) - elif [ -n "$have_fuser" ]; then - # - # Use fuser to free up mount point - # - while read command pid user - do - if [ -z "$pid" ]; then - continue - fi - - if [ $try -eq 1 ]; then - ocf_log warn \ - "killing process $pid ($user $command $mp)" - elif [ $try -eq 3 ]; then - ocf_log crit \ - "Could not clean up mount point $mp" - ret=$FAIL - fi - - if [ $try -gt 1 ]; then - kill -9 $pid - else - kill -TERM $pid - fi - done < <(fuser -vm $mp 2>&1 | \ - grep -v PID | \ - sed 's;^'$mp:';;' | \ - awk '{print $4,$2,$1}' | \ - sort -u -k 1,3) - fi - done - - return $ret -} - -# # startNFSFilesystem # startNFSFilesystem() { @@ -659,15 +552,20 @@ stopNFSFilesystem() { umount_failed=yes - if [ "$force_umount" ]; then - killMountProcesses $mp - fi + if [ "$force_umount" ]; then + if [ $try -eq 1 ]; then + fuser -TERM -kvm "$mp" + else + fuser -kvm "$mp" + fi + fi - if [ $try -ge $max_tries ]; then - done=$YES - else - sleep $sleep_time - let try=try+1 + + if [ $try -ge $max_tries ]; then + done=$YES + else + sleep $sleep_time + let try=try+1 fi ;; *)

1 0

cluster: RHEL56 - rgmanager: Allow exit while waiting for cman
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=fb… Commit: fbc82625c84ca9dfd99f7d8e3e051c53a63bf523 Parent: a8d21b2f0ccb2b9b50f4e729e070fec0fcfd8fee Author: Lon Hohberger <lhh(a)redhat.com> AuthorDate: Mon Mar 30 17:01:45 2009 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 13:57:38 2010 -0400 rgmanager: Allow exit while waiting for cman If cman isn't running, you can now kill rgmanager with SIGINT (e.g. ctrl-C) or SIGTERM to make it quit before initialization is complete. Signed-off-by: Lon Hohberger <lhh(a)redhat.com> Signed-off-by: Fabio M. Di Nitto <fdinitto(a)redhat.com> --- rgmanager/src/daemons/main.c | 17 +++++++++++++---- 1 files changed, 13 insertions(+), 4 deletions(-) diff --git a/rgmanager/src/daemons/main.c b/rgmanager/src/daemons/main.c index 93ecff3..64c32a3 100644 --- a/rgmanager/src/daemons/main.c +++ b/rgmanager/src/daemons/main.c @@ -934,8 +934,8 @@ configure_rgmanager(int ccsfd, int dbg, int *token_secs) } -void -clu_initialize(cman_handle_t *ch) +int +cman_connect(cman_handle_t *ch) { if (!ch) exit(1); @@ -946,6 +946,8 @@ clu_initialize(cman_handle_t *ch) while (!(*ch = cman_init(NULL))) { sleep(1); + if (shutdown_pending) + return 1; } } @@ -960,10 +962,13 @@ clu_initialize(cman_handle_t *ch) while (cman_is_quorate(*ch) == 0) { sleep(1); + if (shutdown_pending) + return 1; } clulog(LOG_NOTICE, "Quorum formed\n"); } + return 0; } @@ -1070,7 +1075,9 @@ main(int argc, char **argv) unblock_signal(SIGSEGV); } - clu_initialize(&clu); + if (cman_connect(&clu) != 0) + goto out; /* Clean exit if sigint/sigterm here */ + if (cman_init_subsys(clu) < 0) { perror("cman_init_subsys"); return -1; @@ -1173,8 +1180,10 @@ main(int argc, char **argv) cleanup(cluster_ctx); out_cleanup: - clulog(LOG_NOTICE, "Shutdown complete, exiting\n"); clu_lock_finished(rgmanager_lsname); + +out: + clulog(LOG_NOTICE, "Shutdown complete, exiting\n"); cman_finish(clu); /*malloc_dump_table(); */ /* Only works if alloc.c us used */

1 0

cluster: RHEL56 - resource-agents: Remove nfs service temp directories
by Lon Hohberger 29 Jul '10

29 Jul '10

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=a8… Commit: a8d21b2f0ccb2b9b50f4e729e070fec0fcfd8fee Parent: 8270fa123369c509fda7ef31db75dfe93d2d3749 Author: Carlos Eduardo Maiolino <cmaiolin(a)redhat.com> AuthorDate: Thu Jul 29 13:45:29 2010 -0400 Committer: Lon Hohberger <lhh(a)redhat.com> CommitterDate: Thu Jul 29 13:48:37 2010 -0400 resource-agents: Remove nfs service temp directories Resolves: rhbz#595455 Signed-off-by: Lon Hohberger <lhh(a)redhat.com> --- rgmanager/src/resources/svclib_nfslock | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/rgmanager/src/resources/svclib_nfslock b/rgmanager/src/resources/svclib_nfslock index a0250b2..2ec7bdd 100644 --- a/rgmanager/src/resources/svclib_nfslock +++ b/rgmanager/src/resources/svclib_nfslock @@ -37,7 +37,7 @@ # nfslock_statd_notify() { - declare tmpdir=$(mktemp -d /tmp/statd-$2.XXXXXX) + declare tmpdir declare nl_dir=$1 declare nl_ip=$2 declare command # Work around bugs in rpc.statd @@ -53,6 +53,8 @@ nfslock_statd_notify() ocf_log debug "No hosts to notify" return 0 fi + + tmpdir=$(mktemp -d /tmp/statd-$2.XXXXXX) # Ok, copy the HA directory to something we can use. mkdir -p $tmpdir/sm

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

cluster-commits July 2010