fence-agents: master - fence_cisco_ucs: protocol prefix of URL was not correctly set with --ssl-(in)secure
by Marek Grác
Gitweb: http://git.fedorahosted.org/git/?p=fence-agents.git;a=commitdiff;h=766e74...
Commit: 766e747471acce46b8e3a5fb77a944008ad5a452
Parent: 0e5893247c7fb881df234d515189f0df6a0b2597
Author: Marek 'marx' Grac <mgrac(a)redhat.com>
AuthorDate: Wed Nov 19 11:04:19 2014 +0100
Committer: Marek 'marx' Grac <mgrac(a)redhat.com>
CommitterDate: Wed Nov 19 11:04:19 2014 +0100
fence_cisco_ucs: protocol prefix of URL was not correctly set with --ssl-(in)secure
Resolves: rhbz#1165591
---
fence/agents/cisco_ucs/fence_cisco_ucs.py | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fence/agents/cisco_ucs/fence_cisco_ucs.py b/fence/agents/cisco_ucs/fence_cisco_ucs.py
index 191f1b1..6288207 100644
--- a/fence/agents/cisco_ucs/fence_cisco_ucs.py
+++ b/fence/agents/cisco_ucs/fence_cisco_ucs.py
@@ -77,7 +77,7 @@ def get_list(conn, options):
def send_command(opt, command, timeout):
## setup correct URL
- if opt.has_key("--ssl"):
+ if "--ssl" in opt or "--ssl-secure" in opt or "--ssl-insecure" in opt:
url = "https:"
else:
url = "http:"
9 years, 4 months
fence-agents: master - fencing: Fix problem with not validating SSL certificates
by Marek Grác
Gitweb: http://git.fedorahosted.org/git/?p=fence-agents.git;a=commitdiff;h=0e5893...
Commit: 0e5893247c7fb881df234d515189f0df6a0b2597
Parent: 01c534d3ecfb794ef7f3b5ea58478740f33c83fb
Author: Marek 'marx' Grac <mgrac(a)redhat.com>
AuthorDate: Mon Nov 10 10:58:54 2014 +0100
Committer: Marek 'marx' Grac <mgrac(a)redhat.com>
CommitterDate: Wed Nov 19 10:58:25 2014 +0100
fencing: Fix problem with not validating SSL certificates
In commit 809b909651118c06d2ab48d7911bbee2e512e478, --ssl-(in)secure options were
added. With --ssl be equal to --ssl-secure this was not implemented properly.
Resolves: rhbz#1162092
---
fence/agents/lib/fencing.py.py | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fence/agents/lib/fencing.py.py b/fence/agents/lib/fencing.py.py
index 8cc2a98..56c84da 100644
--- a/fence/agents/lib/fencing.py.py
+++ b/fence/agents/lib/fencing.py.py
@@ -1036,7 +1036,7 @@ def fence_login(options, re_login_string=r"(login\s*: )|(Login Name: )|(usernam
if options.has_key("--ssl-insecure"):
ssl_opts = "--insecure"
- command = '%s %s %s --insecure --crlf -p %s %s' % \
+ command = '%s %s %s --crlf -p %s %s' % \
(options["--gnutlscli-path"], gnutls_opts, ssl_opts, options["--ipport"], options["--ip"])
try:
conn = fspawn(options, command)
9 years, 4 months
fence-agents: master - fence_mpath: new fence agent for dm-multipath based on mpathpersist
by Marek Grác
Gitweb: http://git.fedorahosted.org/git/?p=fence-agents.git;a=commitdiff;h=01c534...
Commit: 01c534d3ecfb794ef7f3b5ea58478740f33c83fb
Parent: d75b0ced9c203782a444bf51f4926d3c75b0fb5b
Author: Marek 'marx' Grac <mgrac(a)redhat.com>
AuthorDate: Wed Nov 19 10:51:57 2014 +0100
Committer: Marek 'marx' Grac <mgrac(a)redhat.com>
CommitterDate: Wed Nov 19 10:54:01 2014 +0100
fence_mpath: new fence agent for dm-multipath based on mpathpersist
Previously, scenario with multipath and underlying SCSI devices was solved by using
fence_scsi what works correctly but there are some limitation. The most important
is that unfencing has to be done when all paths are available as it is executed only once.
This new fence agent solve this situation properly as most of this situations are solved
by mpathpersist which is part of dm-multipath.
---
configure.ac | 4 +
fence/agents/mpath/Makefile.am | 17 +++
fence/agents/mpath/fence_mpath.py | 244 +++++++++++++++++++++++++++++++++++
make/fencebuild.mk | 3 +
tests/data/metadata/fence_mpath.xml | 86 ++++++++++++
5 files changed, 354 insertions(+), 0 deletions(-)
diff --git a/configure.ac b/configure.ac
index 530d64f..3ce4b06 100644
--- a/configure.ac
+++ b/configure.ac
@@ -175,6 +175,9 @@ AC_PATH_PROG([VGS_PATH], [vgs], [/usr/sbin/vgs])
AC_PATH_PROG([SUDO_PATH], [sudo], [/usr/bin/sudo])
AC_PATH_PROG([SSH_PATH], [ssh], [/usr/bin/ssh])
AC_PATH_PROG([TELNET_PATH], [telnet], [/usr/bin/telnet])
+AC_PATH_PROG([MPATH_PATH], [mpathpersist], [/usr/sbin/mpathpersist])
+AC_PATH_PROG([SUDO_PATH], [sudo], [/usr/bin/sudo])
+
## do subst
AC_SUBST([DEFAULT_CONFIG_DIR])
@@ -295,6 +298,7 @@ AC_CONFIG_FILES([Makefile
fence/agents/lib/Makefile
fence/agents/lpar/Makefile
fence/agents/manual/Makefile
+ fence/agents/mpath/Makefile
fence/agents/netio/Makefile
fence/agents/ovh/Makefile
fence/agents/pve/Makefile
diff --git a/fence/agents/mpath/Makefile.am b/fence/agents/mpath/Makefile.am
new file mode 100644
index 0000000..fd3d8d2
--- /dev/null
+++ b/fence/agents/mpath/Makefile.am
@@ -0,0 +1,17 @@
+MAINTAINERCLEANFILES = Makefile.in
+
+TARGET = fence_mpath
+
+SRC = $(TARGET).py
+
+EXTRA_DIST = $(SRC)
+
+sbin_SCRIPTS = $(TARGET)
+
+man_MANS = $(TARGET).8
+
+FENCE_TEST_ARGS = -k 1
+
+include $(top_srcdir)/make/fencebuild.mk
+include $(top_srcdir)/make/fenceman.mk
+include $(top_srcdir)/make/agentpycheck.mk
diff --git a/fence/agents/mpath/fence_mpath.py b/fence/agents/mpath/fence_mpath.py
new file mode 100644
index 0000000..8a4811c
--- /dev/null
+++ b/fence/agents/mpath/fence_mpath.py
@@ -0,0 +1,244 @@
+#!/usr/bin/python -tt
+
+import sys
+import stat
+import re
+import os
+import logging
+import atexit
+sys.path.append("@FENCEAGENTSLIBDIR@")
+from fencing import fail_usage, run_command, atexit_handler, check_input, process_input, show_docs
+from fencing import fence_action, all_opt, run_delay
+
+#BEGIN_VERSION_GENERATION
+RELEASE_VERSION=""
+REDHAT_COPYRIGHT=""
+BUILD_DATE=""
+#END_VERSION_GENERATION
+
+def get_status(conn, options):
+ del conn
+ status = "off"
+ for dev in options["devices"]:
+ is_block_device(dev)
+ if options["--key"] in get_registration_keys(options, dev):
+ status = "on"
+ else:
+ logging.debug("No registration for key "\
+ + options["--key"] + " on device " + dev + "\n")
+ return status
+
+
+def set_status(conn, options):
+ del conn
+ count = 0
+ if options["--action"] == "on":
+ for dev in options["devices"]:
+ is_block_device(dev)
+
+ register_dev(options, dev)
+ if options["--key"] not in get_registration_keys(options, dev):
+ count += 1
+ logging.debug("Failed to register key "\
+ + options["--key"] + "on device " + dev + "\n")
+ continue
+ dev_write(options, dev)
+
+ if get_reservation_key(options, dev) is None \
+ and not reserve_dev(options, dev) \
+ and get_reservation_key(options, dev) is None:
+ count += 1
+ logging.debug("Failed to create reservation (key="\
+ + options["--key"] + ", device=" + dev + ")\n")
+
+ else:
+ dev_keys = dev_read(options)
+
+ for dev in options["devices"]:
+ is_block_device(dev)
+
+ if options["--key"] in get_registration_keys(options, dev):
+ preempt_abort(options, dev_keys[dev], dev)
+
+ for dev in options["devices"]:
+ if options["--key"] in get_registration_keys(options, dev):
+ count += 1
+ logging.debug("Failed to remove key "\
+ + options["--key"] + " on device " + dev + "\n")
+ continue
+
+ if not get_reservation_key(options, dev):
+ count += 1
+ logging.debug("No reservation exists on device " + dev + "\n")
+ if count:
+ logging.error("Failed to verify " + str(count) + " device(s)")
+ sys.exit(1)
+
+
+#run command, returns dict, ret["err"] = exit code; ret["out"] = output
+def run_cmd(options, cmd):
+ ret = {}
+
+ if options.has_key("--use-sudo"):
+ prefix = options["--sudo-path"] + " "
+ else:
+ prefix = ""
+
+ (ret["err"], ret["out"], _) = run_command(options, prefix + cmd)
+ ret["out"] = "".join([i for i in ret["out"] if i is not None])
+ return ret
+
+
+# check if device exist and is block device
+def is_block_device(dev):
+ if not os.path.exists(dev):
+ fail_usage("Failed: device \"" + dev + "\" does not exist")
+ if not stat.S_ISBLK(os.stat(dev).st_mode):
+ fail_usage("Failed: device \"" + dev + "\" is not a block device")
+
+# cancel registration
+def preempt_abort(options, host, dev):
+ cmd = options["--mpathpersist-path"] + " -o --preempt-abort --prout-type=5 --param-rk=" + host +" --param-sark=" + options["--key"] +"-d " + dev
+ return not bool(run_cmd(options, cmd)["err"])
+
+def register_dev(options, dev):
+ cmd = options["--mpathpersist-path"] + " -o --register --param-sark=" + options["--key"] + " -d " + dev
+ #cmd return code != 0 but registration can be successful
+ return not bool(run_cmd(options, cmd)["err"])
+
+def reserve_dev(options, dev):
+ cmd = options["--mpathpersist-path"] + " -o --reserv --prout-type=5 --param-rk=" + options["--key"] + " -d " + dev
+ return not bool(run_cmd(options, cmd)["err"])
+
+def get_reservation_key(options, dev):
+ cmd = options["--mpathpersist-path"] + " -i -r -d " + dev
+ out = run_cmd(options, cmd)
+ if out["err"]:
+ fail_usage("Cannot get reservation key")
+ match = re.search(r"\s+key\s*=\s*0x(\S+)\s+", out["out"], re.IGNORECASE)
+ return match.group(1) if match else None
+
+def get_registration_keys(options, dev):
+ keys = []
+ cmd = options["--mpathpersist-path"] + " -i -k -d " + dev
+ out = run_cmd(options, cmd)
+ if out["err"]:
+ fail_usage("Cannot get registration keys")
+ for line in out["out"].split("\n"):
+ match = re.search(r"\s+0x(\S+)\s*", line)
+ if match:
+ keys.append(match.group(1))
+ return keys
+
+def dev_write(options, dev):
+ file_path = options["--store-path"] + "/mpath.devices"
+
+ if not os.path.isdir(os.path.dirname(options["--store-path"])):
+ os.makedirs(os.path.dirname(options["--store-path"]))
+
+ try:
+ store_fh = open(file_path, "a+")
+ except IOError:
+ fail_usage("Failed: Cannot open file \""+ file_path + "\"")
+ out = store_fh.read()
+ if not re.search(r"^" + dev + r"\s+", out):
+ store_fh.write(dev + "\t" + options["--key"] + "\n")
+ store_fh.close()
+
+def dev_read(options):
+ dev_key = {}
+ file_path = options["--store-path"] + "/mpath.devices"
+ try:
+ store_fh = open(file_path, "r")
+ except IOError:
+ fail_usage("Failed: Cannot open file \"" + file_path + "\"")
+ # get not empty lines from file
+ for (device, key) in [line.strip().split() for line in store_fh if line.strip()]:
+ dev_key[device] = key
+ store_fh.close()
+ return dev_key
+
+def define_new_opts():
+ all_opt["devices"] = {
+ "getopt" : "d:",
+ "longopt" : "devices",
+ "help" : "-d, --devices=[devices] List of devices to use for current operation",
+ "required" : "0",
+ "shortdesc" : "List of devices to use for current operation. Devices can \
+be comma-separated list of device-mapper multipath devices (eg. /dev/dm-3). \
+Each device must support SCSI-3 persistent reservations.",
+ "order": 1
+ }
+ all_opt["key"] = {
+ "getopt" : "k:",
+ "longopt" : "key",
+ "help" : "-k, --key=[key] Key to use for the current operation",
+ "required" : "1",
+ "shortdesc" : "Key to use for the current operation. This key should be \
+unique to a node and have to be written in /etc/multipath.conf. For the \"on\" action, the key specifies the key use to \
+register the local node. For the \"off\" action, this key specifies the key to \
+be removed from the device(s).",
+ "order": 1
+ }
+ all_opt["mpathpersist_path"] = {
+ "getopt" : "X:",
+ "longopt" : "mpathpersist-path",
+ "help" : "--mpathpersist-path=[path] Path to mpathpersist binary",
+ "required" : "0",
+ "shortdesc" : "Path to mpathpersist binary",
+ "default" : "@MPATH_PATH@",
+ "order": 200
+ }
+ all_opt["store_path"] = {
+ "getopt" : "X:",
+ "longopt" : "store-path",
+ "help" : "--store-path=[path] Path to directory containing cached keys",
+ "required" : "0",
+ "shortdesc" : "Path to directory where fence agent can store information",
+ "default" : "@STORE_PATH@",
+ "order": 200
+ }
+
+def main():
+ atexit.register(atexit_handler)
+
+ device_opt = ["no_login", "no_password", "devices", "key", "sudo", \
+ "fabric_fencing", "on_target", "store_path", "mpathpersist_path"]
+
+ define_new_opts()
+
+ options = check_input(device_opt, process_input(device_opt))
+
+ docs = {}
+ docs["shortdesc"] = "Fence agent for multipath persistent reservation"
+ docs["longdesc"] = "fence_mpath is an I/O fencing agent that uses SCSI-3 \
+persistent reservations to control access multipath devices. Underlying \
+devices must support SCSI-3 persistent reservations (SPC-3 or greater) as \
+well as the \"preempt-and-abort\" subcommand.\nThe fence_mpath agent works by \
+having an unique key for each pair of node and device that has to be set also \
+in /etc/multipath.conf. Once registered, a single node will become the reservation holder \
+by creating a \"write exclusive, registrants only\" reservation on the \
+device(s). The result is that only registered nodes may write to the \
+device(s). When a node failure occurs, the fence_mpath agent will remove the \
+key belonging to the failed node from the device(s). The failed node will no \
+longer be able to write to the device(s). A manual reboot is required."
+ docs["vendorurl"] = "https://www.sourceware.org/dm/"
+ show_docs(options, docs)
+
+ run_delay(options)
+
+ # Input control BEGIN
+ if not "--key" in options:
+ fail_usage("Failed: key is required")
+
+ options["devices"] = options["--devices"].split(",")
+
+ if not options["devices"]:
+ fail_usage("Failed: No devices found")
+ # Input control END
+
+ result = fence_action(None, options, set_status, get_status)
+ sys.exit(result)
+
+if __name__ == "__main__":
+ main()
diff --git a/make/fencebuild.mk b/make/fencebuild.mk
index 3c14928..874ebeb 100644
--- a/make/fencebuild.mk
+++ b/make/fencebuild.mk
@@ -19,6 +19,9 @@ $(TARGET): $(SRC)
-e 's#@''SUDO_PATH@#${SUDO_PATH}#g' \
-e 's#@''SSH_PATH@#${SSH_PATH}#g' \
-e 's#@''TELNET_PATH@#${TELNET_PATH}#g' \
+ -e 's#@''MPATH_PATH@#${MPATH_PATH}#g' \
+ -e 's#@''STORE_PATH@#${CLUSTERVARRUN}#g' \
+ -e 's#@''SUDO_PATH@#${SUDO_PATH}#g' \
> $@
if [ 0 -eq `echo "$(SRC)" | grep fence_ &> /dev/null; echo $$?` ]; then \
diff --git a/tests/data/metadata/fence_mpath.xml b/tests/data/metadata/fence_mpath.xml
new file mode 100644
index 0000000..bd8b141
--- /dev/null
+++ b/tests/data/metadata/fence_mpath.xml
@@ -0,0 +1,86 @@
+<?xml version="1.0" ?>
+<resource-agent name="fence_mpath" shortdesc="Fence agent for multipath persistent reservation" >
+<longdesc>...</longdesc>
+<vendor-url>fence_mpath is an I/O fencing agent that uses SCSI-3 persistent reservations to control access to shared storage devices. These devices must be controlled by dm-multipath and underlying devices must support SCSI-3 persistent reservations (SPC-3 or greater) as well as the "preempt-and-abort" subcommand.
+ The fence_mpath agent works by having unique key for each pair: node and device. These keys have to be manually added to /etc/multipath.conf. Once registered, a single node will become the reservation holder by creating a "write exclusive, registrants only" reservation on the device(s). The result is that only registered nodes may write to the device(s). When a node failure occurs, the fence_scsi agent will remove the key belonging to the failed node from the device(s). The failed node will no longer be able to write to the device(s). A manual reboot is required.</vendor-url>
+<parameters>
+ <parameter name="devices" unique="0" required="0">
+ <getopt mixed="-d, --devices=[devices]" />
+ <content type="string" />
+ <shortdesc lang="en">List of devices to use for current operation. Devices can be comma-separated list of device-mapper multipath devices (eg. /dev/dm-3). Each device must support SCSI-3 persistent reservations.</shortdesc>
+ </parameter>
+ <parameter name="key" unique="0" required="1">
+ <getopt mixed="-k, --key=[key]" />
+ <content type="string" />
+ <shortdesc lang="en">Key to use for the current operation. This key should be unique to a node and have to be written in /etc/multipath.conf. For the "on" action, the key specifies the key use to register the local node. For the "off" action, this key specifies the key to be removed from the device(s).</shortdesc>
+ </parameter>
+ <parameter name="action" unique="0" required="1">
+ <getopt mixed="-o, --action=[action]" />
+ <content type="string" default="off" />
+ <shortdesc lang="en">Fencing Action</shortdesc>
+ </parameter>
+ <parameter name="verbose" unique="0" required="0">
+ <getopt mixed="-v, --verbose" />
+ <content type="boolean" />
+ <shortdesc lang="en">Verbose mode</shortdesc>
+ </parameter>
+ <parameter name="debug" unique="0" required="0">
+ <getopt mixed="-D, --debug-file=[debugfile]" />
+ <content type="string" />
+ <shortdesc lang="en">Write debug information to given file</shortdesc>
+ </parameter>
+ <parameter name="version" unique="0" required="0">
+ <getopt mixed="-V, --version" />
+ <content type="boolean" />
+ <shortdesc lang="en">Display version information and exit</shortdesc>
+ </parameter>
+ <parameter name="help" unique="0" required="0">
+ <getopt mixed="-h, --help" />
+ <content type="boolean" />
+ <shortdesc lang="en">Display help and exit</shortdesc>
+ </parameter>
+ <parameter name="shell_timeout" unique="0" required="0">
+ <getopt mixed="--shell-timeout=[seconds]" />
+ <content type="string" default="3" />
+ <shortdesc lang="en">Wait X seconds for cmd prompt after issuing command</shortdesc>
+ </parameter>
+ <parameter name="power_timeout" unique="0" required="0">
+ <getopt mixed="--power-timeout=[seconds]" />
+ <content type="string" default="20" />
+ <shortdesc lang="en">Test X seconds for status change after ON/OFF</shortdesc>
+ </parameter>
+ <parameter name="power_wait" unique="0" required="0">
+ <getopt mixed="--power-wait=[seconds]" />
+ <content type="string" default="0" />
+ <shortdesc lang="en">Wait X seconds after issuing ON/OFF</shortdesc>
+ </parameter>
+ <parameter name="delay" unique="0" required="0">
+ <getopt mixed="--delay=[seconds]" />
+ <content type="string" default="0" />
+ <shortdesc lang="en">Wait X seconds before fencing is started</shortdesc>
+ </parameter>
+ <parameter name="login_timeout" unique="0" required="0">
+ <getopt mixed="--login-timeout=[seconds]" />
+ <content type="string" default="5" />
+ <shortdesc lang="en">Wait X seconds for cmd prompt after login</shortdesc>
+ </parameter>
+ <parameter name="mpathpersist_path" unique="0" required="0">
+ <getopt mixed="--mpathpersist-path=[path]" />
+ <content type="string" default="/usr/sbin/mpathpersist" />
+ <shortdesc lang="en">Path to sg_persist binary</shortdesc>
+ </parameter>
+ <parameter name="retry_on" unique="0" required="0">
+ <getopt mixed="--retry-on=[attempts]" />
+ <content type="string" default="1" />
+ <shortdesc lang="en">Count of attempts to retry power on</shortdesc>
+ </parameter>
+</parameters>
+<actions>
+ <action name="on" on_target="1" automatic="1"/>
+ <action name="off" />
+ <action name="status" />
+ <action name="list" />
+ <action name="monitor" />
+ <action name="metadata" />
+</actions>
+</resource-agent>
9 years, 4 months
gfs2-utils: RHEL7 - mkfs.gfs2: Revert default resource group size
by Andrew Price
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=33c48a55...
Commit: 33c48a55afa572b86de40364246d749a0e5d44e2
Parent: 86fe75d99f1838410bebaf6f459bfee3333e5756
Author: Andrew Price <anprice(a)redhat.com>
AuthorDate: Thu Nov 13 14:09:08 2014 +0000
Committer: Andrew Price <anprice(a)redhat.com>
CommitterDate: Tue Nov 18 19:04:32 2014 +0000
mkfs.gfs2: Revert default resource group size
Choose a more performant resource group size than the maximum by default
until gfs2 is more tuned for the larger number of bitmap blocks required
for 2GB resource groups.
Resolves: rhbz#1162817
Signed-off-by: Andrew Price <anprice(a)redhat.com>
---
gfs2/mkfs/main_mkfs.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/gfs2/mkfs/main_mkfs.c b/gfs2/mkfs/main_mkfs.c
index e927d82..f4b8fe6 100644
--- a/gfs2/mkfs/main_mkfs.c
+++ b/gfs2/mkfs/main_mkfs.c
@@ -159,7 +159,7 @@ static void opts_init(struct mkfs_opts *opts)
opts->bsize = GFS2_DEFAULT_BSIZE;
opts->jsize = GFS2_DEFAULT_JSIZE;
opts->qcsize = GFS2_DEFAULT_QCSIZE;
- opts->rgsize = GFS2_MAX_RGSIZE;
+ opts->rgsize = GFS2_DEFAULT_RGSIZE;
opts->lockproto = "lock_dlm";
opts->locktable = "";
opts->confirm = 1;
9 years, 4 months
gfs2-utils: RHEL7 - fsck.gfs2: Detect and correct corrupt journals
by Bob Peterson
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=86fe75d9...
Commit: 86fe75d99f1838410bebaf6f459bfee3333e5756
Parent: 34ab6e375e654b5b8b4c4c3c01e12c9f9749a928
Author: Bob Peterson <rpeterso(a)redhat.com>
AuthorDate: Thu Nov 13 09:54:16 2014 -0600
Committer: Bob Peterson <rpeterso(a)redhat.com>
CommitterDate: Tue Nov 18 07:58:59 2014 -0600
fsck.gfs2: Detect and correct corrupt journals
This patch changes fsck.gfs2 so that it can detect corruption in the
journal sequence numbering scheme. If the corruption is within a
certain tolerance (currently 10 sequencing errors) the journal sequence
numbers are fixed and journal is replayed as normal. If there are more
errors than can be tolerated, the problem is detected and the journal
may be cleared.
rhbz#1146160
---
gfs2/fsck/fs_recovery.c | 54 +++++++++++++++++++++++++++++++++++++---------
1 files changed, 43 insertions(+), 11 deletions(-)
diff --git a/gfs2/fsck/fs_recovery.c b/gfs2/fsck/fs_recovery.c
index bc4210d..3d8fe98 100644
--- a/gfs2/fsck/fs_recovery.c
+++ b/gfs2/fsck/fs_recovery.c
@@ -17,6 +17,7 @@
#include "util.h"
#define JOURNAL_NAME_SIZE 16
+#define JOURNAL_SEQ_TOLERANCE 10
unsigned int sd_found_jblocks = 0, sd_replayed_jblocks = 0;
unsigned int sd_found_metablocks = 0, sd_replayed_metablocks = 0;
@@ -138,10 +139,15 @@ static int buf_lo_scan_elements(struct gfs2_inode *ip, unsigned int start,
check_magic = ((struct gfs2_meta_header *)
(bh_ip->b_data))->mh_magic;
check_magic = be32_to_cpu(check_magic);
- if (check_magic != GFS2_MAGIC)
+ if (check_magic != GFS2_MAGIC) {
+ log_err(_("Journal corruption detected at block #"
+ "%lld (0x%llx) for journal+0x%x.\n"),
+ (unsigned long long)blkno, (unsigned long long)blkno,
+ start);
error = -EIO;
- else
+ } else {
bmodified(bh_ip);
+ }
brelse(bh_log);
brelse(bh_ip);
@@ -313,8 +319,11 @@ static int foreach_descriptor(struct gfs2_inode *ip, unsigned int start,
brelse(bh);
continue;
}
- if (error == 1)
+ if (error == 1) {
+ log_err(_("Journal corruption detected at "
+ "journal+0x%x.\n"), start);
error = -EIO;
+ }
bmodified(bh);
brelse(bh);
return error;
@@ -354,10 +363,13 @@ static int foreach_descriptor(struct gfs2_inode *ip, unsigned int start,
}
/**
- * fix_journal_seq_no - Fix log header sequencing problems
+ * check_journal_seq_no - Check and Fix log header sequencing problems
* @ip: the journal incore inode
+ * @fix: if 1, fix the sequence numbers, otherwise just report the problem
+ *
+ * Returns: The number of sequencing errors (hopefully none).
*/
-static int fix_journal_seq_no(struct gfs2_inode *ip)
+static int check_journal_seq_no(struct gfs2_inode *ip, int fix)
{
int error = 0, wrapped = 0;
uint32_t jd_blocks = ip->i_di.di_size / ip->i_sbd->sd_sb.sb_bsize;
@@ -368,6 +380,7 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
uint64_t dblock;
uint32_t extlen;
struct gfs2_buffer_head *bh;
+ int seq_errors = 0;
memset(&lh, 0, sizeof(lh));
for (blk = 0; blk < jd_blocks; blk++) {
@@ -395,6 +408,10 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
(unsigned long long)lowest_seq,
(unsigned long long)highest_seq,
(unsigned long long)prev_seq);
+ if (!fix) {
+ seq_errors++;
+ continue;
+ }
highest_seq++;
lh.lh_sequence = highest_seq;
prev_seq = lh.lh_sequence;
@@ -404,7 +421,9 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
gfs2_log_header_out(&lh, bh);
brelse(bh);
}
- return 0;
+ if (seq_errors && fix)
+ log_err(_("%d sequence errors fixed.\n"), seq_errors);
+ return seq_errors;
}
/**
@@ -456,6 +475,15 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
osi_list_init(&sd_revoke_list);
error = gfs2_find_jhead(ip, &head);
+ if (!error) {
+ error = check_journal_seq_no(ip, 0);
+ if (error > JOURNAL_SEQ_TOLERANCE) {
+ log_err( _("Journal #%d (\"journal%d\") has %d "
+ "sequencing errors; tolerance is %d.\n"),
+ j+1, j, error, JOURNAL_SEQ_TOLERANCE);
+ goto out;
+ }
+ }
if (error) {
if (opts.no) {
log_err( _("Journal #%d (\"journal%d\") is corrupt\n"),j+1, j);
@@ -481,7 +509,7 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
goto out;
}
log_info( _("jid=%u: Repairing journal...\n"), j);
- error = fix_journal_seq_no(ip);
+ error = check_journal_seq_no(ip, 1);
if (error) {
log_err( _("jid=%u: Unable to fix the bad journal.\n"),
j);
@@ -530,8 +558,10 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
for (pass = 0; pass < 2; pass++) {
error = foreach_descriptor(ip, head.lh_tail,
head.lh_blkno, pass);
- if (error)
+ if (error) {
+ log_err(_("Error found during journal replay.\n"));
goto out;
+ }
}
log_info( _("jid=%u: Found %u revoke tags\n"), j, sd_found_revokes);
gfs2_revoke_clean(sdp);
@@ -550,14 +580,16 @@ out:
log_info( _("jid=%u: Done\n"), j);
return 0;
}
- log_info( _("jid=%u: Failed\n"), j);
+ log_err( _("jid=%u: Failed\n"), j);
reinit:
- if (query( _("Do you want to clear the journal instead? (y/n)")))
+ if (query( _("Do you want to clear the journal instead? (y/n)"))) {
error = write_journal(sdp->md.journal[j], sdp->bsize,
sdp->md.journal[j]->i_di.di_size /
sdp->sd_sb.sb_bsize);
- else
+ log_err(_("jid=%u: journal was cleared.\n"), j);
+ } else {
log_err( _("jid=%u: journal not cleared.\n"), j);
+ }
return error;
}
9 years, 4 months
gfs2-utils: master - fsck.gfs2: Detect and correct corrupt journals
by Bob Peterson
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=a292875b...
Commit: a292875bcd3ac77252775c0ca46c930363b718a4
Parent: eb969a757bb420ca8437b0d0be5a893ea8bd8a44
Author: Bob Peterson <rpeterso(a)redhat.com>
AuthorDate: Thu Nov 13 09:54:16 2014 -0600
Committer: Bob Peterson <rpeterso(a)redhat.com>
CommitterDate: Tue Nov 18 07:57:38 2014 -0600
fsck.gfs2: Detect and correct corrupt journals
This patch changes fsck.gfs2 so that it can detect corruption in the
journal sequence numbering scheme. If the corruption is within a
certain tolerance (currently 10 sequencing errors) the journal sequence
numbers are fixed and journal is replayed as normal. If there are more
errors than can be tolerated, the problem is detected and the journal
may be cleared.
---
gfs2/fsck/fs_recovery.c | 54 +++++++++++++++++++++++++++++++++++++---------
1 files changed, 43 insertions(+), 11 deletions(-)
diff --git a/gfs2/fsck/fs_recovery.c b/gfs2/fsck/fs_recovery.c
index bc4210d..3d8fe98 100644
--- a/gfs2/fsck/fs_recovery.c
+++ b/gfs2/fsck/fs_recovery.c
@@ -17,6 +17,7 @@
#include "util.h"
#define JOURNAL_NAME_SIZE 16
+#define JOURNAL_SEQ_TOLERANCE 10
unsigned int sd_found_jblocks = 0, sd_replayed_jblocks = 0;
unsigned int sd_found_metablocks = 0, sd_replayed_metablocks = 0;
@@ -138,10 +139,15 @@ static int buf_lo_scan_elements(struct gfs2_inode *ip, unsigned int start,
check_magic = ((struct gfs2_meta_header *)
(bh_ip->b_data))->mh_magic;
check_magic = be32_to_cpu(check_magic);
- if (check_magic != GFS2_MAGIC)
+ if (check_magic != GFS2_MAGIC) {
+ log_err(_("Journal corruption detected at block #"
+ "%lld (0x%llx) for journal+0x%x.\n"),
+ (unsigned long long)blkno, (unsigned long long)blkno,
+ start);
error = -EIO;
- else
+ } else {
bmodified(bh_ip);
+ }
brelse(bh_log);
brelse(bh_ip);
@@ -313,8 +319,11 @@ static int foreach_descriptor(struct gfs2_inode *ip, unsigned int start,
brelse(bh);
continue;
}
- if (error == 1)
+ if (error == 1) {
+ log_err(_("Journal corruption detected at "
+ "journal+0x%x.\n"), start);
error = -EIO;
+ }
bmodified(bh);
brelse(bh);
return error;
@@ -354,10 +363,13 @@ static int foreach_descriptor(struct gfs2_inode *ip, unsigned int start,
}
/**
- * fix_journal_seq_no - Fix log header sequencing problems
+ * check_journal_seq_no - Check and Fix log header sequencing problems
* @ip: the journal incore inode
+ * @fix: if 1, fix the sequence numbers, otherwise just report the problem
+ *
+ * Returns: The number of sequencing errors (hopefully none).
*/
-static int fix_journal_seq_no(struct gfs2_inode *ip)
+static int check_journal_seq_no(struct gfs2_inode *ip, int fix)
{
int error = 0, wrapped = 0;
uint32_t jd_blocks = ip->i_di.di_size / ip->i_sbd->sd_sb.sb_bsize;
@@ -368,6 +380,7 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
uint64_t dblock;
uint32_t extlen;
struct gfs2_buffer_head *bh;
+ int seq_errors = 0;
memset(&lh, 0, sizeof(lh));
for (blk = 0; blk < jd_blocks; blk++) {
@@ -395,6 +408,10 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
(unsigned long long)lowest_seq,
(unsigned long long)highest_seq,
(unsigned long long)prev_seq);
+ if (!fix) {
+ seq_errors++;
+ continue;
+ }
highest_seq++;
lh.lh_sequence = highest_seq;
prev_seq = lh.lh_sequence;
@@ -404,7 +421,9 @@ static int fix_journal_seq_no(struct gfs2_inode *ip)
gfs2_log_header_out(&lh, bh);
brelse(bh);
}
- return 0;
+ if (seq_errors && fix)
+ log_err(_("%d sequence errors fixed.\n"), seq_errors);
+ return seq_errors;
}
/**
@@ -456,6 +475,15 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
osi_list_init(&sd_revoke_list);
error = gfs2_find_jhead(ip, &head);
+ if (!error) {
+ error = check_journal_seq_no(ip, 0);
+ if (error > JOURNAL_SEQ_TOLERANCE) {
+ log_err( _("Journal #%d (\"journal%d\") has %d "
+ "sequencing errors; tolerance is %d.\n"),
+ j+1, j, error, JOURNAL_SEQ_TOLERANCE);
+ goto out;
+ }
+ }
if (error) {
if (opts.no) {
log_err( _("Journal #%d (\"journal%d\") is corrupt\n"),j+1, j);
@@ -481,7 +509,7 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
goto out;
}
log_info( _("jid=%u: Repairing journal...\n"), j);
- error = fix_journal_seq_no(ip);
+ error = check_journal_seq_no(ip, 1);
if (error) {
log_err( _("jid=%u: Unable to fix the bad journal.\n"),
j);
@@ -530,8 +558,10 @@ static int gfs2_recover_journal(struct gfs2_inode *ip, int j, int preen,
for (pass = 0; pass < 2; pass++) {
error = foreach_descriptor(ip, head.lh_tail,
head.lh_blkno, pass);
- if (error)
+ if (error) {
+ log_err(_("Error found during journal replay.\n"));
goto out;
+ }
}
log_info( _("jid=%u: Found %u revoke tags\n"), j, sd_found_revokes);
gfs2_revoke_clean(sdp);
@@ -550,14 +580,16 @@ out:
log_info( _("jid=%u: Done\n"), j);
return 0;
}
- log_info( _("jid=%u: Failed\n"), j);
+ log_err( _("jid=%u: Failed\n"), j);
reinit:
- if (query( _("Do you want to clear the journal instead? (y/n)")))
+ if (query( _("Do you want to clear the journal instead? (y/n)"))) {
error = write_journal(sdp->md.journal[j], sdp->bsize,
sdp->md.journal[j]->i_di.di_size /
sdp->sd_sb.sb_bsize);
- else
+ log_err(_("jid=%u: journal was cleared.\n"), j);
+ } else {
log_err( _("jid=%u: journal not cleared.\n"), j);
+ }
return error;
}
9 years, 4 months
gfs2-utils: RHEL7 - fsck.gfs2: Improve reporting of pass timings
by Andrew Price
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=34ab6e37...
Commit: 34ab6e375e654b5b8b4c4c3c01e12c9f9749a928
Parent: 9142cd526adfb9739b92419fc1faa0b4730e936d
Author: Andrew Price <anprice(a)redhat.com>
AuthorDate: Thu Nov 13 07:43:01 2014 -0500
Committer: Andrew Price <anprice(a)redhat.com>
CommitterDate: Mon Nov 17 13:39:25 2014 +0000
fsck.gfs2: Improve reporting of pass timings
The days value is not currently reported in the pass timings but the
hours value is still reported modulo 24. Drop the use of gmtime(3) as
it's more appropriate for calendar time operations than elapsed time,
and add a simple duration reporting function which matches the existing
output format.
Resolves: rhbz#1154786
Signed-off-by: Andrew Price <anprice(a)redhat.com>
---
gfs2/fsck/main.c | 48 ++++++++++++++++++++++++++++++++++--------------
1 files changed, 34 insertions(+), 14 deletions(-)
diff --git a/gfs2/fsck/main.c b/gfs2/fsck/main.c
index b25d802..a4af25d 100644
--- a/gfs2/fsck/main.c
+++ b/gfs2/fsck/main.c
@@ -246,18 +246,46 @@ static const struct fsck_pass passes[] = {
{ .name = NULL, }
};
+static void print_pass_duration(const char *name, struct timeval *start)
+{
+ char duration[17] = ""; /* strlen("XXdXXhXXmXX.XXXs") + 1 */
+ struct timeval end, diff;
+ unsigned d, h, m, s;
+ char *p = duration;
+
+ gettimeofday(&end, NULL);
+ timersub(&end, start, &diff);
+
+ s = diff.tv_sec % 60;
+ diff.tv_sec /= 60;
+ m = diff.tv_sec % 60;
+ diff.tv_sec /= 60;
+ h = diff.tv_sec % 24;
+ d = diff.tv_sec / 24;
+
+ if (d)
+ p += snprintf(p, 4, "%ud", d > 99 ? 99U : d);
+ if (h)
+ p += snprintf(p, 4, "%uh", h);
+ if (m)
+ p += snprintf(p, 4, "%um", m);
+
+ snprintf(p, 8, "%u.%03lus", s, diff.tv_usec / 1000);
+ log_notice(_("%s completed in %s\n"), name, duration);
+}
+
static int fsck_pass(const struct fsck_pass *p, struct gfs2_sbd *sdp)
{
int ret;
- struct timeval before, after, diff;
- time_t runtime;
- struct tm *run_tm;
+ struct timeval timer;
if (fsck_abort)
return FSCK_CANCELED;
pass = p->name;
+
log_notice( _("Starting %s\n"), p->name);
- gettimeofday(&before, 0);
+ gettimeofday(&timer, NULL);
+
ret = p->f(sdp);
if (ret)
exit(ret);
@@ -266,16 +294,8 @@ static int fsck_pass(const struct fsck_pass *p, struct gfs2_sbd *sdp)
log_notice( _("%s interrupted \n"), p->name);
return FSCK_CANCELED;
}
- gettimeofday(&after, 0);
- timersub(&after, &before, &diff);
- runtime = (time_t)diff.tv_sec;
- run_tm = gmtime(&runtime);
- log_notice( _("%s completed in "), p->name);
- if (run_tm->tm_hour)
- log_notice("%dh", run_tm->tm_hour);
- if (run_tm->tm_min)
- log_notice("%dm", run_tm->tm_min);
- log_notice("%d.%03lds \n", run_tm->tm_sec, diff.tv_usec / 1000);
+
+ print_pass_duration(p->name, &timer);
return 0;
}
9 years, 4 months
gfs2-utils: master - mkfs.gfs2: Revert default resource group size
by Andrew Price
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=eb969a75...
Commit: eb969a757bb420ca8437b0d0be5a893ea8bd8a44
Parent: 78cca2a4cb19a29284de292646d892d7d5729106
Author: Andrew Price <anprice(a)redhat.com>
AuthorDate: Thu Nov 13 14:09:08 2014 +0000
Committer: Andrew Price <anprice(a)redhat.com>
CommitterDate: Thu Nov 13 14:09:08 2014 +0000
mkfs.gfs2: Revert default resource group size
Choose a more performant resource group size than the maximum by default
until gfs2 is more tuned for the larger number of bitmap blocks required
for 2GB resource groups.
Signed-off-by: Andrew Price <anprice(a)redhat.com>
---
gfs2/mkfs/main_mkfs.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/gfs2/mkfs/main_mkfs.c b/gfs2/mkfs/main_mkfs.c
index b6353dd..2590ec8 100644
--- a/gfs2/mkfs/main_mkfs.c
+++ b/gfs2/mkfs/main_mkfs.c
@@ -147,7 +147,7 @@ static void opts_init(struct mkfs_opts *opts)
opts->bsize = GFS2_DEFAULT_BSIZE;
opts->jsize = GFS2_DEFAULT_JSIZE;
opts->qcsize = GFS2_DEFAULT_QCSIZE;
- opts->rgsize = GFS2_MAX_RGSIZE;
+ opts->rgsize = GFS2_DEFAULT_RGSIZE;
opts->lockproto = "lock_dlm";
opts->locktable = "";
opts->confirm = 1;
9 years, 4 months
gfs2-utils: master - fsck.gfs2: Improve reporting of pass timings
by Andrew Price
Gitweb: http://git.fedorahosted.org/git/?p=gfs2-utils.git;a=commitdiff;h=78cca2a4...
Commit: 78cca2a4cb19a29284de292646d892d7d5729106
Parent: a3647dea1dedb9b4ecf9d7b6de24d990d7f7d915
Author: Andrew Price <anprice(a)redhat.com>
AuthorDate: Thu Nov 13 07:43:01 2014 -0500
Committer: Andrew Price <anprice(a)redhat.com>
CommitterDate: Thu Nov 13 13:59:14 2014 +0000
fsck.gfs2: Improve reporting of pass timings
The days value is not currently reported in the pass timings but the
hours value is still reported modulo 24. Drop the use of gmtime(3) as
it's more appropriate for calendar time operations than elapsed time,
and add a simple duration reporting function which matches the existing
output format.
Signed-off-by: Andrew Price <anprice(a)redhat.com>
---
gfs2/fsck/main.c | 48 ++++++++++++++++++++++++++++++++++--------------
1 files changed, 34 insertions(+), 14 deletions(-)
diff --git a/gfs2/fsck/main.c b/gfs2/fsck/main.c
index b25d802..a4af25d 100644
--- a/gfs2/fsck/main.c
+++ b/gfs2/fsck/main.c
@@ -246,18 +246,46 @@ static const struct fsck_pass passes[] = {
{ .name = NULL, }
};
+static void print_pass_duration(const char *name, struct timeval *start)
+{
+ char duration[17] = ""; /* strlen("XXdXXhXXmXX.XXXs") + 1 */
+ struct timeval end, diff;
+ unsigned d, h, m, s;
+ char *p = duration;
+
+ gettimeofday(&end, NULL);
+ timersub(&end, start, &diff);
+
+ s = diff.tv_sec % 60;
+ diff.tv_sec /= 60;
+ m = diff.tv_sec % 60;
+ diff.tv_sec /= 60;
+ h = diff.tv_sec % 24;
+ d = diff.tv_sec / 24;
+
+ if (d)
+ p += snprintf(p, 4, "%ud", d > 99 ? 99U : d);
+ if (h)
+ p += snprintf(p, 4, "%uh", h);
+ if (m)
+ p += snprintf(p, 4, "%um", m);
+
+ snprintf(p, 8, "%u.%03lus", s, diff.tv_usec / 1000);
+ log_notice(_("%s completed in %s\n"), name, duration);
+}
+
static int fsck_pass(const struct fsck_pass *p, struct gfs2_sbd *sdp)
{
int ret;
- struct timeval before, after, diff;
- time_t runtime;
- struct tm *run_tm;
+ struct timeval timer;
if (fsck_abort)
return FSCK_CANCELED;
pass = p->name;
+
log_notice( _("Starting %s\n"), p->name);
- gettimeofday(&before, 0);
+ gettimeofday(&timer, NULL);
+
ret = p->f(sdp);
if (ret)
exit(ret);
@@ -266,16 +294,8 @@ static int fsck_pass(const struct fsck_pass *p, struct gfs2_sbd *sdp)
log_notice( _("%s interrupted \n"), p->name);
return FSCK_CANCELED;
}
- gettimeofday(&after, 0);
- timersub(&after, &before, &diff);
- runtime = (time_t)diff.tv_sec;
- run_tm = gmtime(&runtime);
- log_notice( _("%s completed in "), p->name);
- if (run_tm->tm_hour)
- log_notice("%dh", run_tm->tm_hour);
- if (run_tm->tm_min)
- log_notice("%dm", run_tm->tm_min);
- log_notice("%d.%03lds \n", run_tm->tm_sec, diff.tv_usec / 1000);
+
+ print_pass_duration(p->name, &timer);
return 0;
}
9 years, 4 months
fence-agents: master - fence_scsi: "order" value of *_path should be same as in library
by Marek Grác
Gitweb: http://git.fedorahosted.org/git/?p=fence-agents.git;a=commitdiff;h=d75b0c...
Commit: d75b0ced9c203782a444bf51f4926d3c75b0fb5b
Parent: 7a04e60e463888fd1a6737cfbb636fbcf8de2d3a
Author: Marek 'marx' Grac <mgrac(a)redhat.com>
AuthorDate: Thu Nov 13 17:52:49 2014 +0100
Committer: Marek 'marx' Grac <mgrac(a)redhat.com>
CommitterDate: Thu Nov 13 17:52:49 2014 +0100
fence_scsi: "order" value of *_path should be same as in library
---
fence/agents/scsi/fence_scsi.py | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/fence/agents/scsi/fence_scsi.py b/fence/agents/scsi/fence_scsi.py
index 9bc5236..894be22 100644
--- a/fence/agents/scsi/fence_scsi.py
+++ b/fence/agents/scsi/fence_scsi.py
@@ -304,7 +304,7 @@ be removed from the device(s).",
"required" : "0",
"shortdesc" : "Path to corosync-cmapctl binary",
"default" : "@COROSYNC_CMAPCTL_PATH@",
- "order": 200
+ "order": 300
}
all_opt["sg_persist_path"] = {
"getopt" : ":",
@@ -313,7 +313,7 @@ be removed from the device(s).",
"required" : "0",
"shortdesc" : "Path to sg_persist binary",
"default" : "@SG_PERSIST_PATH@",
- "order": 200
+ "order": 300
}
all_opt["sg_turs_path"] = {
"getopt" : ":",
@@ -322,7 +322,7 @@ be removed from the device(s).",
"required" : "0",
"shortdesc" : "Path to sg_turs binary",
"default" : "@SG_TURS_PATH@",
- "order": 200
+ "order": 300
}
all_opt["vgs_path"] = {
"getopt" : ":",
@@ -331,7 +331,7 @@ be removed from the device(s).",
"required" : "0",
"shortdesc" : "Path to vgs binary",
"default" : "@VGS_PATH@",
- "order": 200
+ "order": 300
}
9 years, 4 months