diff options
Diffstat (limited to 'extras')
| -rwxr-xr-x | extras/distributed-testing/distributed-test-runner.py | 12 | ||||
| -rw-r--r-- | extras/ganesha/ocf/ganesha_nfsd | 2 | ||||
| -rwxr-xr-x | extras/ganesha/scripts/create-export-ganesha.sh | 1 | ||||
| -rw-r--r-- | extras/ganesha/scripts/ganesha-ha.sh | 93 | ||||
| -rwxr-xr-x | extras/glusterfs-georep-upgrade.py | 77 | ||||
| -rw-r--r-- | extras/group-virt.example | 5 | ||||
| -rwxr-xr-x | extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh | 27 | ||||
| -rwxr-xr-x | extras/hook-scripts/start/post/S31ganesha-start.sh | 2 | ||||
| -rwxr-xr-x | extras/quota/quota_fsck.py | 8 | ||||
| -rwxr-xr-x | extras/snap_scheduler/gcron.py | 7 | ||||
| -rwxr-xr-x | extras/snap_scheduler/snap_scheduler.py | 4 | ||||
| -rwxr-xr-x | extras/statedumpparse.rb | 151 | ||||
| -rw-r--r-- | extras/systemd/glusterd.service.in | 5 |
13 files changed, 275 insertions, 119 deletions
diff --git a/extras/distributed-testing/distributed-test-runner.py b/extras/distributed-testing/distributed-test-runner.py index 7bfb6c9652a..5a07e2feab1 100755 --- a/extras/distributed-testing/distributed-test-runner.py +++ b/extras/distributed-testing/distributed-test-runner.py @@ -383,14 +383,17 @@ class Handlers: return self.shell.call("make install") == 0 @synchronized - def prove(self, id, test, timeout, valgrind=False, asan_noleaks=True): + def prove(self, id, test, timeout, valgrind="no", asan_noleaks=True): assert id == self.client_id self.shell.cd(self.gluster_root) env = "DEBUG=1 " - if valgrind: + if valgrind == "memcheck" or valgrind == "yes": cmd = "valgrind" cmd += " --tool=memcheck --leak-check=full --track-origins=yes" cmd += " --show-leak-kinds=all -v prove -v" + elif valgrind == "drd": + cmd = "valgrind" + cmd += " --tool=drd -v prove -v" elif asan_noleaks: cmd = "prove -v" env += "ASAN_OPTIONS=detect_leaks=0 " @@ -827,8 +830,9 @@ parser.add_argument("--port", help="server port to listen", type=int, default=DEFAULT_PORT) # test role parser.add_argument("--tester", help="start tester", action="store_true") -parser.add_argument("--valgrind", help="run tests under valgrind", - action="store_true") +parser.add_argument("--valgrind[=memcheck,drd]", + help="run tests with valgrind tool 'memcheck' or 'drd'", + default="no") parser.add_argument("--asan", help="test with asan enabled", action="store_true") parser.add_argument("--asan-noleaks", help="test with asan but no mem leaks", diff --git a/extras/ganesha/ocf/ganesha_nfsd b/extras/ganesha/ocf/ganesha_nfsd index 93fc8be5136..f91e8b6b8f7 100644 --- a/extras/ganesha/ocf/ganesha_nfsd +++ b/extras/ganesha/ocf/ganesha_nfsd @@ -36,7 +36,7 @@ else . ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs fi -OCF_RESKEY_ha_vol_mnt_default="/var/run/gluster/shared_storage" +OCF_RESKEY_ha_vol_mnt_default="/run/gluster/shared_storage" : ${OCF_RESKEY_ha_vol_mnt=${OCF_RESKEY_ha_vol_mnt_default}} ganesha_meta_data() { diff --git a/extras/ganesha/scripts/create-export-ganesha.sh b/extras/ganesha/scripts/create-export-ganesha.sh index 1ffba427457..3040e8138b0 100755 --- a/extras/ganesha/scripts/create-export-ganesha.sh +++ b/extras/ganesha/scripts/create-export-ganesha.sh @@ -64,6 +64,7 @@ echo " Pseudo=\"/$VOL\";" echo ' Protocols = "3", "4" ;' echo ' Transports = "UDP","TCP";' echo ' SecType = "sys";' +echo ' Security_Label = False;' echo " }" } if [ "$OPTION" = "on" ]; diff --git a/extras/ganesha/scripts/ganesha-ha.sh b/extras/ganesha/scripts/ganesha-ha.sh index 32af1cab10e..9790a719e10 100644 --- a/extras/ganesha/scripts/ganesha-ha.sh +++ b/extras/ganesha/scripts/ganesha-ha.sh @@ -24,11 +24,16 @@ GANESHA_HA_SH=$(realpath $0) HA_NUM_SERVERS=0 HA_SERVERS="" HA_VOL_NAME="gluster_shared_storage" -HA_VOL_MNT="/var/run/gluster/shared_storage" +HA_VOL_MNT="/run/gluster/shared_storage" HA_CONFDIR=$HA_VOL_MNT"/nfs-ganesha" SERVICE_MAN="DISTRO_NOT_FOUND" -RHEL6_PCS_CNAME_OPTION="--name" +# rhel, fedora id, version +ID="" +VERSION_ID="" + +PCS9OR10_PCS_CNAME_OPTION="" +PCS9OR10_PCS_CLONE_OPTION="clone" SECRET_PEM="/var/lib/glusterd/nfs/secret.pem" # UNBLOCK RA uses shared_storage which may become unavailable @@ -101,9 +106,9 @@ determine_service_manager () { then SERVICE_MAN="/sbin/service" fi - if [ "${SERVICE_MAN}" == "DISTRO_NOT_FOUND" ] + if [[ "${SERVICE_MAN}X" == "DISTRO_NOT_FOUNDX" ]] then - echo "Service manager not recognized, exiting" + logger "Service manager not recognized, exiting" exit 1 fi } @@ -114,7 +119,7 @@ manage_service () local new_node=${2} local option= - if [ "${action}" == "start" ]; then + if [[ "${action}" == "start" ]]; then option="yes" else option="no" @@ -122,7 +127,7 @@ manage_service () ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \ ${SECRET_PEM} root@${new_node} "${GANESHA_HA_SH} --setup-ganesha-conf-files $HA_CONFDIR $option" - if [ "${SERVICE_MAN}" == "/bin/systemctl" ] + if [[ "${SERVICE_MAN}" == "/bin/systemctl" ]] then ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \ ${SECRET_PEM} root@${new_node} "${SERVICE_MAN} ${action} nfs-ganesha" @@ -140,7 +145,7 @@ check_cluster_exists() if [ -e /var/run/corosync.pid ]; then cluster_name=$(pcs status | grep "Cluster name:" | cut -d ' ' -f 3) - if [ ${cluster_name} -a ${cluster_name} = ${name} ]; then + if [[ "${cluster_name}X" == "${name}X" ]]; then logger "$name already exists, exiting" exit 0 fi @@ -155,7 +160,7 @@ determine_servers() local tmp_ifs=${IFS} local ha_servers="" - if [ "X${cmd}X" != "XsetupX" -a "X${cmd}X" != "XstatusX" ]; then + if [ "${cmd}X" != "setupX" -a "${cmd}X" != "statusX" ]; then ha_servers=$(pcs status | grep "Online:" | grep -o '\[.*\]' | sed -e 's/\[//' | sed -e 's/\]//') IFS=$' ' for server in ${ha_servers} ; do @@ -193,15 +198,21 @@ setup_cluster() logger "setting up cluster ${name} with the following ${servers}" - pcs cluster auth ${servers} - # pcs cluster setup --name ${name} ${servers} - pcs cluster setup ${RHEL6_PCS_CNAME_OPTION} ${name} --enable --transport udpu ${servers} + # pcs cluster setup --force ${PCS9OR10_PCS_CNAME_OPTION} ${name} ${servers} + pcs cluster setup --force ${PCS9OR10_PCS_CNAME_OPTION} ${name} --enable ${servers} if [ $? -ne 0 ]; then - logger "pcs cluster setup ${RHEL6_PCS_CNAME_OPTION} ${name} --enable --transport udpu ${servers} failed" + logger "pcs cluster setup ${PCS9OR10_PCS_CNAME_OPTION} ${name} --enable ${servers} failed, shutting down ganesha and bailing out" #set up failed stop all ganesha process and clean up symlinks in cluster stop_ganesha_all "${servers}" exit 1; fi + + # pcs cluster auth ${servers} + pcs cluster auth + if [ $? -ne 0 ]; then + logger "pcs cluster auth failed" + fi + pcs cluster start --all if [ $? -ne 0 ]; then logger "pcs cluster start failed" @@ -217,7 +228,7 @@ setup_cluster() done unclean=$(pcs status | grep -u "UNCLEAN") - while [[ "${unclean}X" = "UNCLEANX" ]]; do + while [[ "${unclean}X" == "UNCLEANX" ]]; do sleep 1 unclean=$(pcs status | grep -u "UNCLEAN") done @@ -244,7 +255,7 @@ setup_finalize_ha() local stopped="" stopped=$(pcs status | grep -u "Stopped") - while [[ "${stopped}X" = "StoppedX" ]]; do + while [[ "${stopped}X" == "StoppedX" ]]; do sleep 1 stopped=$(pcs status | grep -u "Stopped") done @@ -265,7 +276,7 @@ refresh_config () if [ -e ${SECRET_PEM} ]; then while [[ ${3} ]]; do current_host=`echo ${3} | cut -d "." -f 1` - if [ ${short_host} != ${current_host} ]; then + if [[ ${short_host} != ${current_host} ]]; then output=$(ssh -oPasswordAuthentication=no \ -oStrictHostKeyChecking=no -i ${SECRET_PEM} root@${current_host} \ "dbus-send --print-reply --system --dest=org.ganesha.nfsd \ @@ -398,7 +409,7 @@ wrap_create_virt_ip_constraints() # the result is "node2 node3 node4"; for node2, "node3 node4 node1" # and so on. while [[ ${1} ]]; do - if [ "${1}" = "${primary}" ]; then + if [[ ${1} == ${primary} ]]; then shift while [[ ${1} ]]; do tail=${tail}" "${1} @@ -429,15 +440,15 @@ setup_create_resources() local cibfile=$(mktemp -u) # fixup /var/lib/nfs - logger "pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone" - pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone + logger "pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION}" + pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION} if [ $? -ne 0 ]; then - logger "warning: pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone failed" + logger "warning: pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION} failed" fi - pcs resource create nfs-mon ocf:heartbeat:ganesha_mon --clone + pcs resource create nfs-mon ocf:heartbeat:ganesha_mon ${PCS9OR10_PCS_CLONE_OPTION} if [ $? -ne 0 ]; then - logger "warning: pcs resource create nfs-mon ocf:heartbeat:ganesha_mon --clone failed" + logger "warning: pcs resource create nfs-mon ocf:heartbeat:ganesha_mon ${PCS9OR10_PCS_CLONE_OPTION} failed" fi # see comment in (/usr/lib/ocf/resource.d/heartbeat/ganesha_grace @@ -445,9 +456,9 @@ setup_create_resources() # ganesha-active crm_attribute sleep 5 - pcs resource create nfs-grace ocf:heartbeat:ganesha_grace --clone notify=true + pcs resource create nfs-grace ocf:heartbeat:ganesha_grace ${PCS9OR10_PCS_CLONE_OPTION} notify=true if [ $? -ne 0 ]; then - logger "warning: pcs resource create nfs-grace ocf:heartbeat:ganesha_grace --clone failed" + logger "warning: pcs resource create nfs-grace ocf:heartbeat:ganesha_grace ${PCS9OR10_PCS_CLONE_OPTION} failed" fi pcs constraint location nfs-grace-clone rule score=-INFINITY grace-active ne 1 @@ -616,7 +627,7 @@ addnode_recreate_resources() --after ${add_node}-nfs_block if [ $? -ne 0 ]; then logger "warning pcs resource create ${add_node}-cluster_ip-1 ocf:heartbeat:IPaddr \ - ip=${add_vip} cidr_netmask=32 op monitor interval=15s failed" + ip=${add_vip} cidr_netmask=32 op monitor interval=15s failed" fi pcs -f ${cibfile} constraint order nfs-grace-clone then ${add_node}-cluster_ip-1 @@ -780,7 +791,7 @@ setup_state_volume() touch ${mnt}/nfs-ganesha/${dirname}/nfs/statd/state fi for server in ${HA_SERVERS} ; do - if [ ${server} != ${dirname} ]; then + if [[ ${server} != ${dirname} ]]; then ln -s ${mnt}/nfs-ganesha/${server}/nfs/ganesha ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/${server} ln -s ${mnt}/nfs-ganesha/${server}/nfs/statd ${mnt}/nfs-ganesha/${dirname}/nfs/statd/${server} fi @@ -794,7 +805,7 @@ setup_state_volume() enable_pacemaker() { while [[ ${1} ]]; do - if [ "${SERVICE_MAN}" == "/usr/bin/systemctl" ]; then + if [[ "${SERVICE_MAN}" == "/bin/systemctl" ]]; then ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \ ${SECRET_PEM} root@${1} "${SERVICE_MAN} enable pacemaker" else @@ -892,7 +903,7 @@ delnode_state_volume() rm -rf ${mnt}/nfs-ganesha/${dirname} for server in ${HA_SERVERS} ; do - if [[ "${server}" != "${dirname}" ]]; then + if [[ ${server} != ${dirname} ]]; then rm -f ${mnt}/nfs-ganesha/${server}/nfs/ganesha/${dirname} rm -f ${mnt}/nfs-ganesha/${server}/nfs/statd/${dirname} fi @@ -908,8 +919,9 @@ status() local index=1 local nodes - # change tabs to spaces, strip leading spaces - pcs status | sed -e "s/\t/ /g" -e "s/^[ ]*//" > ${scratch} + # change tabs to spaces, strip leading spaces, including any + # new '*' at the beginning of a line introduced in pcs-0.10.x + pcs status | sed -e "s/\t/ /g" -e "s/^[ ]*\*//" -e "s/^[ ]*//" > ${scratch} nodes[0]=${1}; shift @@ -924,7 +936,7 @@ status() done # print the nodes that are expected to be online - grep -E "^Online:" ${scratch} + grep -E "Online:" ${scratch} echo @@ -963,7 +975,7 @@ status() create_ganesha_conf_file() { - if [ $1 == "yes" ]; + if [[ "$1" == "yes" ]]; then if [ -e $GANESHA_CONF ]; then @@ -1012,6 +1024,13 @@ main() semanage boolean -m gluster_use_execmem --on fi + local osid="" + + osid=$(grep ^ID= /etc/os-release) + eval $(echo ${osid} | grep -F ID=) + osid=$(grep ^VERSION_ID= /etc/os-release) + eval $(echo ${osid} | grep -F VERSION_ID=) + HA_CONFDIR=${1%/}; shift local ha_conf=${HA_CONFDIR}/ganesha-ha.conf local node="" @@ -1032,7 +1051,17 @@ main() determine_servers "setup" - if [ "X${HA_NUM_SERVERS}X" != "X1X" ]; then + # Fedora 29+ and rhel/centos 8 has PCS-0.10.x + # default is pcs-0.10.x options but check for + # rhel/centos 7 (pcs-0.9.x) and adjust accordingly + if [[ ! ${ID} =~ {rhel,centos} ]]; then + if [[ ${VERSION_ID} == 7.* ]]; then + PCS9OR10_PCS_CNAME_OPTION="--name" + PCS9OR10_PCS_CLONE_OPTION="--clone" + fi + fi + + if [[ "${HA_NUM_SERVERS}X" != "1X" ]]; then determine_service_manager diff --git a/extras/glusterfs-georep-upgrade.py b/extras/glusterfs-georep-upgrade.py new file mode 100755 index 00000000000..634576058d6 --- /dev/null +++ b/extras/glusterfs-georep-upgrade.py @@ -0,0 +1,77 @@ +#!/usr/bin/python3 +""" + +Copyright (c) 2020 Red Hat, Inc. <http://www.redhat.com> +This file is part of GlusterFS. + +This file is licensed to you under your choice of the GNU Lesser +General Public License, version 3 or any later version (LGPLv3 or +later), or the GNU General Public License, version 2 (GPLv2), in all +cases as published by the Free Software Foundation. + +""" + +import argparse +import errno +import os, sys +import shutil +from datetime import datetime + +def find_htime_path(brick_path): + dirs = [] + htime_dir = os.path.join(brick_path, '.glusterfs/changelogs/htime') + for file in os.listdir(htime_dir): + if os.path.isfile(os.path.join(htime_dir,file)) and file.startswith("HTIME"): + dirs.append(os.path.join(htime_dir, file)) + else: + raise FileNotFoundError("%s unavailable" % (os.path.join(htime_dir, file))) + return dirs + +def modify_htime_file(brick_path): + htime_file_path_list = find_htime_path(brick_path) + + for htime_file_path in htime_file_path_list: + changelog_path = os.path.join(brick_path, '.glusterfs/changelogs') + temp_htime_path = os.path.join(changelog_path, 'htime/temp_htime_file') + with open(htime_file_path, 'r') as htime_file, open(temp_htime_path, 'w') as temp_htime_file: + #extract epoch times from htime file + paths = htime_file.read().split("\x00") + + for pth in paths: + epoch_no = pth.split(".")[-1] + changelog = os.path.basename(pth) + #convert epoch time to year, month and day + if epoch_no != '': + date=(datetime.fromtimestamp(float(int(epoch_no))).strftime("%Y/%m/%d")) + #update paths in temp htime file + temp_htime_file.write("%s/%s/%s\x00" % (changelog_path, date, changelog)) + #create directory in the format year/month/days + path = os.path.join(changelog_path, date) + + if changelog.startswith("CHANGELOG."): + try: + os.makedirs(path, mode = 0o600); + except OSError as exc: + if exc.errno == errno.EEXIST: + pass + else: + raise + + #copy existing changelogs to new directory structure, delete old changelog files + shutil.copyfile(pth, os.path.join(path, changelog)) + os.remove(pth) + + #rename temp_htime_file with htime file + os.rename(htime_file_path, os.path.join('%s.bak'%htime_file_path)) + os.rename(temp_htime_path, htime_file_path) + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument('brick_path', help="This upgrade script, which is to be run on\ + server side, takes brick path as the argument, \ + updates paths inside htime file and alters the directory structure \ + above the changelog files inorder to support new optimised format \ + of the directory structure as per \ + https://review.gluster.org/#/c/glusterfs/+/23733/") + args = parser.parse_args() + modify_htime_file(args.brick_path) diff --git a/extras/group-virt.example b/extras/group-virt.example index 6a66f977b2f..cc37c98a25c 100644 --- a/extras/group-virt.example +++ b/extras/group-virt.example @@ -17,3 +17,8 @@ cluster.choose-local=off client.event-threads=4 server.event-threads=4 performance.client-io-threads=on +network.ping-timeout=20 +server.tcp-user-timeout=20 +server.keepalive-time=10 +server.keepalive-interval=2 +server.keepalive-count=5 diff --git a/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh b/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh index 885ed03ad5b..1f2564b44ff 100755 --- a/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh +++ b/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh @@ -79,9 +79,9 @@ done if [ "$option" == "disable" ]; then # Unmount the volume on all the nodes - umount /var/run/gluster/shared_storage - cat /etc/fstab | grep -v "gluster_shared_storage /var/run/gluster/shared_storage/" > /var/run/gluster/fstab.tmp - mv /var/run/gluster/fstab.tmp /etc/fstab + umount /run/gluster/shared_storage + cat /etc/fstab | grep -v "gluster_shared_storage /run/gluster/shared_storage/" > /run/gluster/fstab.tmp + mv /run/gluster/fstab.tmp /etc/fstab fi if [ "$is_originator" == 1 ]; then @@ -104,8 +104,15 @@ function check_volume_status() echo $status } -mount_cmd="mount -t glusterfs $local_node_hostname:/gluster_shared_storage \ - /var/run/gluster/shared_storage" +key=`echo $5 | cut -d '=' -f 1` +val=`echo $5 | cut -d '=' -f 2` +if [ "$key" == "transport.address-family" ]; then + mount_cmd="mount -t glusterfs -o xlator-option=transport.address-family=inet6 \ + $local_node_hostname:/gluster_shared_storage /run/gluster/shared_storage" +else + mount_cmd="mount -t glusterfs $local_node_hostname:/gluster_shared_storage \ + /run/gluster/shared_storage" +fi if [ "$option" == "enable" ]; then retry=0; @@ -120,10 +127,10 @@ if [ "$option" == "enable" ]; then status=$(check_volume_status) done # Mount the volume on all the nodes - umount /var/run/gluster/shared_storage - mkdir -p /var/run/gluster/shared_storage + umount /run/gluster/shared_storage + mkdir -p /run/gluster/shared_storage $mount_cmd - cp /etc/fstab /var/run/gluster/fstab.tmp - echo "$local_node_hostname:/gluster_shared_storage /var/run/gluster/shared_storage/ glusterfs defaults 0 0" >> /var/run/gluster/fstab.tmp - mv /var/run/gluster/fstab.tmp /etc/fstab + cp /etc/fstab /run/gluster/fstab.tmp + echo "$local_node_hostname:/gluster_shared_storage /run/gluster/shared_storage/ glusterfs defaults 0 0" >> /run/gluster/fstab.tmp + mv /run/gluster/fstab.tmp /etc/fstab fi diff --git a/extras/hook-scripts/start/post/S31ganesha-start.sh b/extras/hook-scripts/start/post/S31ganesha-start.sh index 90ba6bc73a5..7ad6f23ad06 100755 --- a/extras/hook-scripts/start/post/S31ganesha-start.sh +++ b/extras/hook-scripts/start/post/S31ganesha-start.sh @@ -4,7 +4,7 @@ OPTSPEC="volname:,gd-workdir:" VOL= declare -i EXPORT_ID ganesha_key="ganesha.enable" -GANESHA_DIR="/var/run/gluster/shared_storage/nfs-ganesha" +GANESHA_DIR="/run/gluster/shared_storage/nfs-ganesha" CONF1="$GANESHA_DIR/ganesha.conf" GLUSTERD_WORKDIR= diff --git a/extras/quota/quota_fsck.py b/extras/quota/quota_fsck.py index 174f2a2a44c..e62f7fc52a3 100755 --- a/extras/quota/quota_fsck.py +++ b/extras/quota/quota_fsck.py @@ -156,12 +156,10 @@ def get_quota_xattr_brick(dpath): xattr_dict = {} xattr_dict['parents'] = {} - for xattr in pairs: + for xattr in pairs[1:]: + xattr = xattr.decode("utf-8") xattr_key = xattr.split("=")[0] - if re.search("# file:", xattr_key): - # skip the file comment - continue - elif xattr_key is "": + if xattr_key == "": # skip any empty lines continue elif not re.search("quota", xattr_key): diff --git a/extras/snap_scheduler/gcron.py b/extras/snap_scheduler/gcron.py index 1127be0e976..0e4df77d481 100755 --- a/extras/snap_scheduler/gcron.py +++ b/extras/snap_scheduler/gcron.py @@ -19,10 +19,10 @@ import logging.handlers import fcntl -GCRON_TASKS = "/var/run/gluster/shared_storage/snaps/glusterfs_snap_cron_tasks" +GCRON_TASKS = "/run/gluster/shared_storage/snaps/glusterfs_snap_cron_tasks" GCRON_CROND_TASK = "/etc/cron.d/glusterfs_snap_cron_tasks" GCRON_RELOAD_FLAG = "/var/run/gluster/crond_task_reload_flag" -LOCK_FILE_DIR = "/var/run/gluster/shared_storage/snaps/lock_files/" +LOCK_FILE_DIR = "/run/gluster/shared_storage/snaps/lock_files/" log = logging.getLogger("gcron-logger") start_time = 0.0 @@ -38,7 +38,8 @@ def initLogger(script_name): sh.setFormatter(formatter) process = subprocess.Popen(["gluster", "--print-logdir"], - stdout=subprocess.PIPE) + stdout=subprocess.PIPE, + universal_newlines=True) out, err = process.communicate() if process.returncode == 0: logfile = os.path.join(out.strip(), script_name[:-3]+".log") diff --git a/extras/snap_scheduler/snap_scheduler.py b/extras/snap_scheduler/snap_scheduler.py index a66c5e3d5ce..e8fcc449a9b 100755 --- a/extras/snap_scheduler/snap_scheduler.py +++ b/extras/snap_scheduler/snap_scheduler.py @@ -67,7 +67,7 @@ except ImportError: SCRIPT_NAME = "snap_scheduler" scheduler_enabled = False log = logging.getLogger(SCRIPT_NAME) -SHARED_STORAGE_DIR="/var/run/gluster/shared_storage" +SHARED_STORAGE_DIR="/run/gluster/shared_storage" GCRON_DISABLED = SHARED_STORAGE_DIR+"/snaps/gcron_disabled" GCRON_ENABLED = SHARED_STORAGE_DIR+"/snaps/gcron_enabled" GCRON_TASKS = SHARED_STORAGE_DIR+"/snaps/glusterfs_snap_cron_tasks" @@ -149,7 +149,7 @@ def initLogger(): sh.setFormatter(formatter) process = subprocess.Popen(["gluster", "--print-logdir"], - stdout=subprocess.PIPE) + stdout=subprocess.PIPE, universal_newlines=True) logfile = os.path.join(process.stdout.read()[:-1], SCRIPT_NAME + ".log") fh = logging.FileHandler(logfile) diff --git a/extras/statedumpparse.rb b/extras/statedumpparse.rb index 872f284fcb8..1aff43377db 100755 --- a/extras/statedumpparse.rb +++ b/extras/statedumpparse.rb @@ -17,25 +17,35 @@ end # client.c:client_fd_lk_ctx_dump uses a six-dash record separator. ARRSEP = /^(-{5,6}=-{5,6})?$/ HEAD = /^\[(.*)\]$/ +INPUT_FORMATS = %w[statedump json] format = 'json' +input_format = 'statedump' tz = '+0000' memstat_select,memstat_reject = //,/\Z./ -human = false OptionParser.new do |op| op.banner << " [<] <STATEDUMP>" - op.on("-f", "--format=F", "json/yaml/memstat") { |s| format = s } + op.on("-f", "--format=F", "json/yaml/memstat(-[plain|human|json])") { |s| format = s } + op.on("--input-format=F", INPUT_FORMATS.join(?/)) { |s| input_format = s } op.on("--timezone=T", "time zone to apply to zoneless timestamps [default UTC]") { |s| tz = s } - op.on("--memstat-select=RX", "memstat: select memory types maxtching RX") { |s| + op.on("--memstat-select=RX", "memstat: select memory types matching RX") { |s| memstat_select = Regexp.new s } - op.on("--memstat-reject=RX", "memstat: reject memory types maxtching RX") { |s| + op.on("--memstat-reject=RX", "memstat: reject memory types matching RX") { |s| memstat_reject = Regexp.new s } - op.on("--memstat-human", "memstat: human readable") { human = true } end.parse! + +if format =~ /\Amemstat(?:-(.*))?/ + memstat_type = $1 || 'plain' + unless %w[plain human json].include? memstat_type + raise "unknown memstat type #{memstat_type.dump}" + end + format = 'memstat' +end + repr, logsep = case format when 'yaml' require 'yaml' @@ -50,8 +60,11 @@ else end formatter = proc { |e| puts repr.call(e) } -d1 = {} +INPUT_FORMATS.include? input_format or raise "unkwown input format '#{input_format}'" + +dumpinfo = {} +# parse a statedump entry elem_cbk = proc { |s,&cbk| arraylike = false s.grep(/\S/).empty? and next @@ -74,15 +87,8 @@ elem_cbk = proc { |s,&cbk| body.reject(&:empty?).map { |e| ea = e.map { |l| k,v = l.split("=",2) - [k, begin - Integer v - rescue ArgumentError - begin - Float v - rescue ArgumentError - v - end - end] + m = /\A(0|-?[1-9]\d*)(\.\d+)?\Z/.match v + [k, m ? (m[2] ? Float(v) : Integer(v)) : v] } begin ea.to_h @@ -93,83 +99,110 @@ elem_cbk = proc { |s,&cbk| } if body - cbk.call [head, arraylike ? body : body[0]] + cbk.call [head, arraylike ? body : (body.empty? ? {} : body[0])] else STDERR.puts ["WARNING: failed to parse record:", repr.call(s)].join(logsep) end } +# aggregator routine aggr = case format when 'memstat' - mh = {} + meminfo = {} + # commit memory-related entries to meminfo proc { |k,r| case k when /memusage/ - (mh["GF_MALLOC"]||={})[k] ||= r["size"] if k =~ memstat_select and k !~ memstat_reject + (meminfo["GF_MALLOC"]||={})[k] ||= r["size"] if k =~ memstat_select and k !~ memstat_reject when "mempool" r.each {|e| kk = "mempool:#{e['pool-name']}" - (mh["mempool"]||={})[kk] ||= e["size"] if kk =~ memstat_select and kk !~ memstat_reject + (meminfo["mempool"]||={})[kk] ||= e["size"] if kk =~ memstat_select and kk !~ memstat_reject } end } else + # just format data, don't actually aggregate anything proc { |pair| formatter.call pair } end -acc = [] -$<.each { |l| - l = l.strip - if l =~ /^(DUMP-(?:START|END)-TIME):\s+(.*)/ - d1["_meta"]||={} - (d1["_meta"]["date"]||={})[$1] = Time.parse([$2, tz].join " ") - next - end +# processing the data +case input_format +when 'statedump' + acc = [] + $<.each { |l| + l = l.strip + if l =~ /^(DUMP-(?:START|END)-TIME):\s+(.*)/ + dumpinfo["_meta"]||={} + (dumpinfo["_meta"]["date"]||={})[$1] = Time.parse([$2, tz].join " ") + next + end - if l =~ HEAD - elem_cbk.call(acc, &aggr) - acc = [l] - next - end + if l =~ HEAD + elem_cbk.call(acc, &aggr) + acc = [l] + next + end - acc << l -} -elem_cbk.call(acc, &aggr) + acc << l + } + elem_cbk.call(acc, &aggr) +when 'json' + $<.each { |l| + r = JSON.load l + case r + when Array + aggr[r] + when Hash + dumpinfo.merge! r + end + } +end +# final actions: output aggregated data case format when 'memstat' - hr = proc { |n| - qa = %w[B kB MB GB] - q = ((1...qa.size).find {|i| n < (1 << i*10)} || qa.size) - 1 - [n*100 / (1 << q*10) / 100.0, qa[q]].join - } - - ma = mh.values.map(&:to_a).inject(:+) - totals = mh.map { |coll,h| [coll, h.values.inject(:+)] }.to_h + ma = meminfo.values.map(&:to_a).inject(:+) + totals = meminfo.map { |coll,h| [coll, h.values.inject(:+)] }.to_h tt = ma.transpose[1].inject(:+) - templ = "%{val} %{key}" - tft = proc { |t| t } - nft = if human - nw = [ma.transpose[1], totals.values, tt].flatten.map{|n| hr[n].size}.max - proc { |n| - hn = hr[n] - " " * (nw - hn.size) + hn + summary_sep,showm = case memstat_type + when 'json' + ["", proc { |k,v| puts({type: k, value: v}.to_json) }] + when 'plain', 'human' + # human-friendly number representation + hr = proc { |n| + qa = %w[B kB MB GB] + q = ((1...qa.size).find {|i| n < (1 << i*10)} || qa.size) - 1 + "%.2f%s" % [n.to_f / (1 << q*10), qa[q]] } + + templ = "%{val} %{key}" + tft = proc { |t| t } + nft = if memstat_type == 'human' + nw = [ma.transpose[1], totals.values, tt].flatten.map{|n| hr[n].size}.max + proc { |n| + hn = hr[n] + " " * (nw - hn.size) + hn + } + else + nw = tt.to_s.size + proc { |n| "%#{nw}d" % n } + end + ## Alternative template, key first: + # templ = "%{key} %{val}" + # tw = ma.transpose[0].map(&:size).max + # tft = proc { |t| t + " " * [tw - t.size, 0].max } + # nft = (memstat_type == 'human') ? hr : proc { |n| n } + ["\n", proc { |k,v| puts templ % {key: tft[k], val: nft[v]} }] else - nw = tt.to_s.size - proc { |n| "%#{nw}d" % n } + raise 'this should be impossible' end - # templ = "%{key} %{val}" - # tw = ma.transpose[0].map(&:size).max - # tft = proc { |t| t + " " * [tw - t.size, 0].max } - # nft = human ? hr : proc { |n| n } - showm = proc { |k,v| puts templ % {key: tft[k], val: nft[v]} } ma.sort_by { |k,v| v }.each(&showm) - puts + print summary_sep totals.each { |coll,t| showm.call "Total #{coll}", t } showm.call "TOTAL", tt else - formatter.call d1 + formatter.call dumpinfo end diff --git a/extras/systemd/glusterd.service.in b/extras/systemd/glusterd.service.in index 32542d9f78b..abb0d82911f 100644 --- a/extras/systemd/glusterd.service.in +++ b/extras/systemd/glusterd.service.in @@ -1,6 +1,8 @@ [Unit] Description=GlusterFS, a clustered file-system server Documentation=man:glusterd(8) +StartLimitBurst=6 +StartLimitIntervalSec=3600 Requires=@RPCBIND_SERVICE@ After=network.target @RPCBIND_SERVICE@ Before=network-online.target @@ -10,7 +12,7 @@ Type=forking PIDFile=@localstatedir@/run/glusterd.pid LimitNOFILE=65536 Environment="LOG_LEVEL=INFO" -EnvironmentFile=-@sysconfdir@/sysconfig/glusterd +EnvironmentFile=-@SYSCONF_DIR@/sysconfig/glusterd ExecStart=@prefix@/sbin/glusterd -p @localstatedir@/run/glusterd.pid --log-level $LOG_LEVEL $GLUSTERD_OPTIONS KillMode=process TimeoutSec=300 @@ -18,7 +20,6 @@ SuccessExitStatus=15 Restart=on-abnormal RestartSec=60 StartLimitBurst=6 -StartLimitIntervalSec=3600 StartLimitInterval=3600 [Install] |