lvm2-commits September 2013

lvm2-commits@lists.fedorahosted.org

6 participants
107 discussions

master - cleanup: monitoring more readable
by Zdenek Kabelac 23 Sep '13

23 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=861a3b2f1940cd36… Commit: 861a3b2f1940cd3667cc353c762ff3d1ef75c37b Parent: cafde608909b87aedeb76f0eda1397f7bf98fea4 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Sep 5 11:32:42 2013 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Sep 23 11:35:15 2013 +0200 cleanup: monitoring more readable Put continue path into one code segment. --- lib/activate/activate.c | 18 +++++++----------- 1 files changed, 7 insertions(+), 11 deletions(-) diff --git a/lib/activate/activate.c b/lib/activate/activate.c index 7b535c6..ecd35b2 100644 --- a/lib/activate/activate.c +++ b/lib/activate/activate.c @@ -1514,22 +1514,18 @@ int monitor_dev_for_events(struct cmd_context *cmd, struct logical_volume *lv, !monitor_dev_for_events(cmd, seg->metadata_lv, NULL, monitor)) r = 0; - if (!seg_monitored(seg) || (seg->status & PVMOVE)) + if (!seg_monitored(seg) || + (seg->status & PVMOVE) || + !seg->segtype->ops->target_monitored) /* doesn't support registration */ continue; - monitor_fn = NULL; - - /* Check monitoring status */ - if (seg->segtype->ops->target_monitored) - monitored = seg->segtype->ops->target_monitored(seg, &pending); - else - continue; /* segtype doesn't support registration */ + monitored = seg->segtype->ops->target_monitored(seg, &pending); - /* - * FIXME: We should really try again if pending - */ + /* FIXME: We should really try again if pending */ monitored = (pending) ? 0 : monitored; + monitor_fn = NULL; + if (monitor) { if (monitored) log_verbose("%s/%s already monitored.", lv->vg->name, lv->name);

1 0

master - test: Blacklist fedora-19 kernel for RAID4/5/6 dev replace tests
by Jonathan Brassow 20 Sep '13

20 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=cafde608909b87ae… Commit: cafde608909b87aedeb76f0eda1397f7bf98fea4 Parent: bd75844024ad4a2e66c7743c50e189a00c233284 Author: Jonathan Brassow <jbrassow(a)redhat.com> AuthorDate: Fri Sep 20 11:33:29 2013 -0500 Committer: Jonathan Brassow <jbrassow(a)redhat.com> CommitterDate: Fri Sep 20 11:33:29 2013 -0500 test: Blacklist fedora-19 kernel for RAID4/5/6 dev replace tests A know issue with kmem_cach is causing failures while testing RAID 4/5/6 device replacement. Blacklist the offending kernel so that these tests are not performed there. --- test/lib/aux.sh | 33 +++++++++++++++++++++++++++++++++ test/shell/lvconvert-raid.sh | 2 ++ 2 files changed, 35 insertions(+), 0 deletions(-) diff --git a/test/lib/aux.sh b/test/lib/aux.sh index 6194492..7939949 100644 --- a/test/lib/aux.sh +++ b/test/lib/aux.sh @@ -615,6 +615,39 @@ skip_if_mirror_recovery_broken() { if test `uname -r` = 3.3.4-5.fc17.x86_64; then skip; fi } +skip_if_raid456_replace_broken() { +# The way kmem_cache aliasing is done in the kernel is broken. +# It causes RAID 4/5/6 tests to fail. +# +# The problem with kmem_cache* is this: +# *) Assume CONFIG_SLUB is set +# 1) kmem_cache_create(name="foo-a") +# - creates new kmem_cache structure +# 2) kmem_cache_create(name="foo-b") +# - If identical cache characteristics, it will be merged with the previously +# created cache associated with "foo-a". The cache's refcount will be +# incremented and an alias will be created via sysfs_slab_alias(). +# 3) kmem_cache_destroy(<ptr>) +# - Attempting to destroy cache associated with "foo-a", but instead the +# refcount is simply decremented. I don't even think the sysfs aliases are +# ever removed... +# 4) kmem_cache_create(name="foo-a") +# - This FAILS because kmem_cache_sanity_check colides with the existing +# name ("foo-a") associated with the non-removed cache. +# +# This is a problem for RAID (specifically dm-raid) because the name used +# for the kmem_cache_create is ("raid%d-%p", level, mddev). If the cache +# persists for long enough, the memory address of an old mddev will be +# reused for a new mddev - causing an identical formulation of the cache +# name. Even though kmem_cache_destory had long ago been used to delete +# the old cache, the merging of caches has cause the name and cache of that +# old instance to be preserved and causes a colision (and thus failure) in +# kmem_cache_create(). I see this regularly in testing the following +# kernels: + if test `uname -r` = 3.10.11-200.fc19.i686; then skip; fi + if test `uname -r` = 3.10.11-200.fc19.x86_64; then skip; fi +} + udev_wait() { pgrep udev >/dev/null || return 0 which udevadm >/dev/null || return 0 diff --git a/test/shell/lvconvert-raid.sh b/test/shell/lvconvert-raid.sh index 4c08386..6762db2 100644 --- a/test/shell/lvconvert-raid.sh +++ b/test/shell/lvconvert-raid.sh @@ -206,6 +206,8 @@ for i in {1..3}; do lvremove -ff $vg done +aux skip_if_raid456_replace_broken + # RAID 4/5/6 (can replace up to 'parity' devices) for i in 4 5 6; do lvcreate --type raid$i -i 3 -l 3 -n $lv1 $vg

1 0

master - release 2.02.101
by Alasdair Kergon 20 Sep '13

20 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=bd75844024ad4a2e… Commit: bd75844024ad4a2e66c7743c50e189a00c233284 Parent: f393a5c15631f883883df41e85b3209b33005cd6 Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Fri Sep 20 13:56:29 2013 +0100 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Fri Sep 20 13:56:29 2013 +0100 release 2.02.101 112 files changed, 4131 insertions(+), 1312 deletions(-) --- VERSION | 2 +- VERSION_DM | 2 +- WHATS_NEW | 6 ++++-- WHATS_NEW_DM | 4 ++-- 4 files changed, 8 insertions(+), 6 deletions(-) diff --git a/VERSION b/VERSION index 4aa9105..8c1db53 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -2.02.101(2)-git (2013-08-13) +2.02.101(2)-git (2013-09-20) diff --git a/VERSION_DM b/VERSION_DM index 2fe4e25..10b2320 100644 --- a/VERSION_DM +++ b/VERSION_DM @@ -1 +1 @@ -1.02.80-git (2013-08-13) +1.02.80-git (2013-09-20) diff --git a/WHATS_NEW b/WHATS_NEW index 0b53a9e..e969620 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,7 @@ -Version 2.02.101 - -=================================== +Version 2.02.101 - 20th September 2013 +====================================== + Fix 3-thread clvmd deadlock triggered by cleanup on EOF from client. + Remove VG from lvmetad before restoring it with vgcfgrestore. Use strtoull instead of strtol in _get_int_arg. Add devtypes report command to display built-in recognised block device types. Fix CC Makefile override which had reverted to using built-in value. (2.02.75) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 51c771c..176234c 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,5 @@ -Version 1.02.80 - -================================== +Version 1.02.80 - 20th September 2013 +===================================== Detect invalid sector supplied to 'dmsetup message'. Free any previously-set string if a dm_task_set_* function is called again. Do not allow passing empty new name for dmsetup rename.

1 0

master - udev: remove unused line in 69-dm-lvm-metad.rules
by Peter Rajnoha 20 Sep '13

20 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f393a5c15631f883… Commit: f393a5c15631f883883df41e85b3209b33005cd6 Parent: 2c41c8b88610bff6a3352b93f6b5c70a3744c817 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Sep 20 09:42:27 2013 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Sep 20 09:45:34 2013 +0200 udev: remove unused line in 69-dm-lvm-metad.rules The explicit check for *_raid_member is not actually needed as this gets filtered out by the ENV{ID_FS_TYPE}!="LVM2_member|LVM1_member" rule. --- udev/69-dm-lvm-metad.rules.in | 6 +----- 1 files changed, 1 insertions(+), 5 deletions(-) diff --git a/udev/69-dm-lvm-metad.rules.in b/udev/69-dm-lvm-metad.rules.in index 6363a79..ba43396 100644 --- a/udev/69-dm-lvm-metad.rules.in +++ b/udev/69-dm-lvm-metad.rules.in @@ -27,6 +27,7 @@ ENV{LVM_PV_GONE}=="1", GOTO="lvm_scan" # Only process devices already marked as a PV - this requires blkid to be called before. ENV{ID_FS_TYPE}!="LVM2_member|LVM1_member", GOTO="lvm_end" +ENV{DM_MULTIPATH_DEVICE_PATH}=="1", GOTO="lvm_end" # Inform lvmetad about any PV that is gone. ACTION=="remove", GOTO="lvm_scan" @@ -76,11 +77,6 @@ LABEL="lvm_scan" # MD | | X | X* | | # loop | | X | X* | | # other | X | | X | | X - -# Skip device that is a multipath or RAID component -ENV{DM_MULTIPATH_DEVICE_PATH}=="1", GOTO="lvm_end" -ENV{ID_FS_TYPE}=="*_raid_member", GOTO="lvm_end" - RUN+="(LVM_EXEC)/lvm pvscan --background --cache --activate ay --major $major --minor $minor", ENV{LVM_SCANNED}="1" LABEL="lvm_end"

1 0

master - RAID: Don't allow syncaction changes on non-RAID LVs
by Jonathan Brassow 20 Sep '13

20 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=2c41c8b88610bff6… Commit: 2c41c8b88610bff6a3352b93f6b5c70a3744c817 Parent: 25bed9968155f43ef0b8832383ef711e7ae1685d Author: Jonathan Brassow <jbrassow(a)redhat.com> AuthorDate: Thu Sep 19 22:33:01 2013 -0500 Committer: Jonathan Brassow <jbrassow(a)redhat.com> CommitterDate: Thu Sep 19 22:33:01 2013 -0500 RAID: Don't allow syncaction changes on non-RAID LVs Don't allow syncaction or other RAID-type messages on non-RAID logical volumes. --- lib/activate/activate.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/lib/activate/activate.c b/lib/activate/activate.c index e5033c6..7b535c6 100644 --- a/lib/activate/activate.c +++ b/lib/activate/activate.c @@ -896,6 +896,12 @@ int lv_raid_message(const struct logical_volume *lv, const char *msg) struct dev_manager *dm; struct dm_status_raid *status; + if (!seg_is_raid(first_seg(lv))) { + log_error("%s/%s must be a RAID logical volume to" + " perform this action.", lv->vg->name, lv->name); + return 0; + } + if (!lv_is_active_locally(lv)) { log_error("Unable to send message to an inactive logical volume."); return 0;

1 0

master - clvmd: Avoid a 3-way deadlock in dead-client cleanup.
by Petr Rockai 18 Sep '13

18 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=25bed9968155f43e… Commit: 25bed9968155f43ef0b8832383ef711e7ae1685d Parent: 3df50d822b8064c5536968556c466c50a9a3ace5 Author: Petr Rockai <prockai(a)redhat.com> AuthorDate: Mon Sep 9 00:01:44 2013 +0200 Committer: Petr Rockai <prockai(a)redhat.com> CommitterDate: Wed Sep 18 21:17:48 2013 +0200 clvmd: Avoid a 3-way deadlock in dead-client cleanup. --- daemons/clvmd/clvmd.c | 177 ++++++++++++++++++++++++++----------------------- daemons/clvmd/clvmd.h | 1 + 2 files changed, 96 insertions(+), 82 deletions(-) diff --git a/daemons/clvmd/clvmd.c b/daemons/clvmd/clvmd.c index a52b39d..bed8e68 100644 --- a/daemons/clvmd/clvmd.c +++ b/daemons/clvmd/clvmd.c @@ -119,6 +119,7 @@ static void *pre_and_post_thread(void *arg); static int send_message(void *buf, int msglen, const char *csid, int fd, const char *errtext); static int read_from_local_sock(struct local_client *thisfd); +static int cleanup_zombie(struct local_client *thisfd); static int process_local_command(struct clvm_header *msg, int msglen, struct local_client *client, unsigned short xid); @@ -698,6 +699,7 @@ static int local_rendezvous_callback(struct local_client *thisfd, char *buf, newfd->bits.localsock.sent_out = FALSE; newfd->bits.localsock.threadid = 0; newfd->bits.localsock.finished = 0; + newfd->bits.localsock.cleanup_needed = 0; newfd->bits.localsock.pipe_client = NULL; newfd->bits.localsock.private = NULL; newfd->bits.localsock.all_success = 1; @@ -889,7 +891,7 @@ static void main_loop(int local_sock, int cmd_timeout) for (thisfd = &local_client_head; thisfd != NULL; thisfd = thisfd->next) { - if (thisfd->removeme) { + if (thisfd->removeme && !cleanup_zombie(thisfd)) { struct local_client *free_fd; lastfd->next = thisfd->next; free_fd = thisfd; @@ -916,7 +918,6 @@ static void main_loop(int local_sock, int cmd_timeout) /* Got error or EOF: Remove it from the list safely */ if (ret <= 0) { - struct local_client *free_fd; int type = thisfd->type; /* If the cluster socket shuts down, so do we */ @@ -926,12 +927,7 @@ static void main_loop(int local_sock, int cmd_timeout) DEBUGLOG("ret == %d, errno = %d. removing client\n", ret, errno); - lastfd->next = thisfd->next; - free_fd = thisfd; - safe_close(&(free_fd->fd)); - - /* Queue cleanup, this also frees the client struct */ - add_to_lvmqueue(free_fd, NULL, 0, NULL); + thisfd->removeme = 1; break; } @@ -1174,6 +1170,95 @@ static void dump_message(char *buf, int len) } } +static int cleanup_zombie(struct local_client *thisfd) +{ + int *status; + + if (thisfd->type != LOCAL_SOCK) + return 0; + + if (!thisfd->bits.localsock.cleanup_needed) + return 0; + + DEBUGLOG("EOF on local socket: inprogress=%d\n", + thisfd->bits.localsock.in_progress); + + thisfd->bits.localsock.finished = 1; + + /* If the client went away in mid command then tidy up */ + if (thisfd->bits.localsock.in_progress) { + pthread_kill(thisfd->bits.localsock.threadid, SIGUSR2); + if (pthread_mutex_trylock(&thisfd->bits.localsock.mutex)) + goto bail; + thisfd->bits.localsock.state = POST_COMMAND; + pthread_cond_signal(&thisfd->bits.localsock.cond); + pthread_mutex_unlock(&thisfd->bits.localsock.mutex); + + /* Free any unsent buffers */ + free_reply(thisfd); + } + + /* Kill the subthread & free resources */ + if (thisfd->bits.localsock.threadid) { + DEBUGLOG("Waiting for child thread\n"); + pthread_mutex_lock(&thisfd->bits.localsock.mutex); + thisfd->bits.localsock.state = PRE_COMMAND; + pthread_cond_signal(&thisfd->bits.localsock.cond); + pthread_mutex_unlock(&thisfd->bits.localsock.mutex); + + if ((errno = pthread_join(thisfd->bits.localsock.threadid, + (void **) &status))) + log_sys_error("pthread_join", ""); + + DEBUGLOG("Joined child thread\n"); + + thisfd->bits.localsock.threadid = 0; + pthread_cond_destroy(&thisfd->bits.localsock.cond); + pthread_mutex_destroy(&thisfd->bits.localsock.mutex); + + /* Remove the pipe client */ + if (thisfd->bits.localsock.pipe_client != NULL) { + struct local_client *newfd; + struct local_client *lastfd = NULL; + struct local_client *free_fd = NULL; + + (void) close(thisfd->bits.localsock.pipe_client->fd); /* Close pipe */ + (void) close(thisfd->bits.localsock.pipe); + + /* Remove pipe client */ + for (newfd = &local_client_head; newfd != NULL; + newfd = newfd->next) { + if (thisfd->bits.localsock. + pipe_client == newfd) { + thisfd->bits.localsock. + pipe_client = NULL; + + lastfd->next = newfd->next; + free_fd = newfd; + newfd->next = lastfd; + free(free_fd); + break; + } + lastfd = newfd; + } + } + } + + /* Free the command buffer */ + free(thisfd->bits.localsock.cmd); + + /* Clear out the cross-link */ + if (thisfd->bits.localsock.pipe_client != NULL) + thisfd->bits.localsock.pipe_client->bits.pipe.client = + NULL; + + safe_close(&(thisfd->fd)); + thisfd->bits.localsock.cleanup_needed = 0; + return 0; +bail: + return 1; +} + /* Called when we have a read from the local socket. was in the main loop but it's grown up and is a big girl now */ static int read_from_local_sock(struct local_client *thisfd) @@ -1204,80 +1289,8 @@ static int read_from_local_sock(struct local_client *thisfd) /* EOF or error on socket */ if (len <= 0) { - int *status; - - DEBUGLOG("EOF on local socket: inprogress=%d\n", - thisfd->bits.localsock.in_progress); - - thisfd->bits.localsock.finished = 1; - - /* If the client went away in mid command then tidy up */ - if (thisfd->bits.localsock.in_progress) { - pthread_kill(thisfd->bits.localsock.threadid, SIGUSR2); - pthread_mutex_lock(&thisfd->bits.localsock.mutex); - thisfd->bits.localsock.state = POST_COMMAND; - pthread_cond_signal(&thisfd->bits.localsock.cond); - pthread_mutex_unlock(&thisfd->bits.localsock.mutex); - - /* Free any unsent buffers */ - free_reply(thisfd); - } - - /* Kill the subthread & free resources */ - if (thisfd->bits.localsock.threadid) { - DEBUGLOG("Waiting for child thread\n"); - pthread_mutex_lock(&thisfd->bits.localsock.mutex); - thisfd->bits.localsock.state = PRE_COMMAND; - pthread_cond_signal(&thisfd->bits.localsock.cond); - pthread_mutex_unlock(&thisfd->bits.localsock.mutex); - - if ((errno = pthread_join(thisfd->bits.localsock.threadid, - (void **) &status))) - log_sys_error("pthread_join", ""); - - DEBUGLOG("Joined child thread\n"); - - thisfd->bits.localsock.threadid = 0; - pthread_cond_destroy(&thisfd->bits.localsock.cond); - pthread_mutex_destroy(&thisfd->bits.localsock.mutex); - - /* Remove the pipe client */ - if (thisfd->bits.localsock.pipe_client != NULL) { - struct local_client *newfd; - struct local_client *lastfd = NULL; - struct local_client *free_fd = NULL; - - (void) close(thisfd->bits.localsock.pipe_client->fd); /* Close pipe */ - (void) close(thisfd->bits.localsock.pipe); - - /* Remove pipe client */ - for (newfd = &local_client_head; newfd != NULL; - newfd = newfd->next) { - if (thisfd->bits.localsock. - pipe_client == newfd) { - thisfd->bits.localsock. - pipe_client = NULL; - - lastfd->next = newfd->next; - free_fd = newfd; - newfd->next = lastfd; - free(free_fd); - break; - } - lastfd = newfd; - } - } - } - - /* Free the command buffer */ - free(thisfd->bits.localsock.cmd); - - /* Clear out the cross-link */ - if (thisfd->bits.localsock.pipe_client != NULL) - thisfd->bits.localsock.pipe_client->bits.pipe.client = - NULL; - - safe_close(&(thisfd->fd)); + thisfd->bits.localsock.cleanup_needed = 1; + cleanup_zombie(thisfd); /* we ignore errors here */ return 0; } else { int comms_pipe[2]; diff --git a/daemons/clvmd/clvmd.h b/daemons/clvmd/clvmd.h index fb0589d..0f837dd 100644 --- a/daemons/clvmd/clvmd.h +++ b/daemons/clvmd/clvmd.h @@ -53,6 +53,7 @@ struct localsock_bits { int finished; /* Flag to tell subthread to exit */ int all_success; /* Set to 0 if any node (or the pre_command) failed */ + int cleanup_needed; /* helper for cleanup_zombie */ struct local_client *pipe_client; pthread_t threadid; enum { PRE_COMMAND, POST_COMMAND, QUIT } state;

1 0

master - vgconvert: Do not call lvmetad_vg_remove (path shared with vgcfgbackup).
by Petr Rockai 18 Sep '13

18 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3df50d822b8064c5… Commit: 3df50d822b8064c5536968556c466c50a9a3ace5 Parent: 054cf25b5f6a34de8886ba6293deee2b24720902 Author: Petr Rockai <prockai(a)redhat.com> AuthorDate: Wed Sep 18 12:53:11 2013 +0200 Committer: Petr Rockai <prockai(a)redhat.com> CommitterDate: Wed Sep 18 12:53:11 2013 +0200 vgconvert: Do not call lvmetad_vg_remove (path shared with vgcfgbackup). --- lib/format_text/archiver.c | 6 +++--- lib/format_text/archiver.h | 2 +- tools/vgconvert.c | 2 +- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/lib/format_text/archiver.c b/lib/format_text/archiver.c index 752e93c..e3e7777 100644 --- a/lib/format_text/archiver.c +++ b/lib/format_text/archiver.c @@ -308,7 +308,7 @@ struct volume_group *backup_read_vg(struct cmd_context *cmd, } /* ORPHAN and VG locks held before calling this */ -int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg) +int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg, int drop_lvmetad) { struct pv_list *pvl; struct format_instance *fid; @@ -351,7 +351,7 @@ int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg) if (!vg_write(vg)) return_0; - if (lvmetad_active()) { + if (drop_lvmetad && lvmetad_active()) { struct volume_group *vg_lvmetad = lvmetad_vg_lookup(cmd, vg->name, NULL); if (vg_lvmetad) { lvmetad_vg_remove(vg_lvmetad); @@ -397,7 +397,7 @@ int backup_restore_from_file(struct cmd_context *cmd, const char *vg_name, missing_pvs = vg_missing_pv_count(vg); if (missing_pvs == 0) - r = backup_restore_vg(cmd, vg); + r = backup_restore_vg(cmd, vg, 1); else log_error("Cannot restore Volume Group %s with %i PVs " "marked as missing.", vg->name, missing_pvs); diff --git a/lib/format_text/archiver.h b/lib/format_text/archiver.h index ddff687..bc756d5 100644 --- a/lib/format_text/archiver.h +++ b/lib/format_text/archiver.h @@ -51,7 +51,7 @@ int backup_remove(struct cmd_context *cmd, const char *vg_name); struct volume_group *backup_read_vg(struct cmd_context *cmd, const char *vg_name, const char *file); -int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg); +int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg, int drop_lvmetad); int backup_restore_from_file(struct cmd_context *cmd, const char *vg_name, const char *file, int force); int backup_restore(struct cmd_context *cmd, const char *vg_name, int force); diff --git a/tools/vgconvert.c b/tools/vgconvert.c index fbb1b4e..86e3097 100644 --- a/tools/vgconvert.c +++ b/tools/vgconvert.c @@ -184,7 +184,7 @@ static int vgconvert_single(struct cmd_context *cmd, const char *vg_name, log_verbose("Writing metadata for VG %s using format %s", vg_name, cmd->fmt->name); - if (!backup_restore_vg(cmd, vg)) { + if (!backup_restore_vg(cmd, vg, 0)) { log_error("Conversion failed for volume group %s.", vg_name); log_error("Use pvcreate and vgcfgrestore to repair from " "archived metadata.");

1 0

master - vgcfgrestore: Remove VG rom lvmetad later, to better deal with errors.
by Petr Rockai 18 Sep '13

18 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=054cf25b5f6a34de… Commit: 054cf25b5f6a34de8886ba6293deee2b24720902 Parent: a6af611ae127de202be9e172099c6efe9ed41d7e Author: Petr Rockai <prockai(a)redhat.com> AuthorDate: Wed Sep 18 11:22:49 2013 +0200 Committer: Petr Rockai <prockai(a)redhat.com> CommitterDate: Wed Sep 18 11:24:58 2013 +0200 vgcfgrestore: Remove VG rom lvmetad later, to better deal with errors. --- lib/format_text/archiver.c | 14 +++++++++++++- tools/vgcfgrestore.c | 8 -------- 2 files changed, 13 insertions(+), 9 deletions(-) diff --git a/lib/format_text/archiver.c b/lib/format_text/archiver.c index b6b9fb3..752e93c 100644 --- a/lib/format_text/archiver.c +++ b/lib/format_text/archiver.c @@ -18,6 +18,7 @@ #include "format-text.h" #include "lvm-string.h" #include "lvmcache.h" +#include "lvmetad.h" #include "toolcontext.h" #include "locking.h" @@ -347,7 +348,18 @@ int backup_restore_vg(struct cmd_context *cmd, struct volume_group *vg) vg->extent_size = tmp; } - if (!vg_write(vg) || !vg_commit(vg)) + if (!vg_write(vg)) + return_0; + + if (lvmetad_active()) { + struct volume_group *vg_lvmetad = lvmetad_vg_lookup(cmd, vg->name, NULL); + if (vg_lvmetad) { + lvmetad_vg_remove(vg_lvmetad); + release_vg(vg_lvmetad); + } + } + + if (!vg_commit(vg)) return_0; return 1; diff --git a/tools/vgcfgrestore.c b/tools/vgcfgrestore.c index f560836..36396c3 100644 --- a/tools/vgcfgrestore.c +++ b/tools/vgcfgrestore.c @@ -57,14 +57,6 @@ int vgcfgrestore(struct cmd_context *cmd, int argc, char **argv) return ECMD_FAILED; } - if (lvmetad_active()) { - struct volume_group *vg = lvmetad_vg_lookup(cmd, vg_name, NULL); - if (vg) { - lvmetad_vg_remove(vg); - release_vg(vg); - } - } - cmd->handles_unknown_segments = 1; if (!(arg_count(cmd, file_ARG) ?

1 0

master - vgcfgrestore: Remove the VG from lvmetad before overwriting it.
by Petr Rockai 18 Sep '13

18 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a6af611ae127de20… Commit: a6af611ae127de202be9e172099c6efe9ed41d7e Parent: c2dd0a832a8f06471f86e8244884c4dc902cc9f7 Author: Petr Rockai <prockai(a)redhat.com> AuthorDate: Wed Sep 18 10:34:38 2013 +0200 Committer: Petr Rockai <prockai(a)redhat.com> CommitterDate: Wed Sep 18 10:37:29 2013 +0200 vgcfgrestore: Remove the VG from lvmetad before overwriting it. --- tools/vgcfgrestore.c | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/tools/vgcfgrestore.c b/tools/vgcfgrestore.c index 36396c3..f560836 100644 --- a/tools/vgcfgrestore.c +++ b/tools/vgcfgrestore.c @@ -57,6 +57,14 @@ int vgcfgrestore(struct cmd_context *cmd, int argc, char **argv) return ECMD_FAILED; } + if (lvmetad_active()) { + struct volume_group *vg = lvmetad_vg_lookup(cmd, vg_name, NULL); + if (vg) { + lvmetad_vg_remove(vg); + release_vg(vg); + } + } + cmd->handles_unknown_segments = 1; if (!(arg_count(cmd, file_ARG) ?

1 0

master - man: lvs man page was not accurate on the volume health bit of lv_attr
by Jonathan Brassow 18 Sep '13

18 Sep '13

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c2dd0a832a8f0647… Commit: c2dd0a832a8f06471f86e8244884c4dc902cc9f7 Parent: 68f841fcda29ee0af177b0299b52842afcf1bae2 Author: Jonathan Brassow <jbrassow(a)redhat.com> AuthorDate: Tue Sep 17 23:03:57 2013 -0500 Committer: Jonathan Brassow <jbrassow(a)redhat.com> CommitterDate: Tue Sep 17 23:03:57 2013 -0500 man: lvs man page was not accurate on the volume health bit of lv_attr The 'm'ismatches flag only shows after a "check" scrubbing operation - not after a "check" or "repair" as indicated by the man page. --- man/lvs.8.in | 11 +++++------ 1 files changed, 5 insertions(+), 6 deletions(-) diff --git a/man/lvs.8.in b/man/lvs.8.in index 8f97dc2..32cb788 100644 --- a/man/lvs.8.in +++ b/man/lvs.8.in @@ -173,12 +173,11 @@ more of the Physical Volumes this RAID Logical Volume uses had suffered a write error. The write error could be due to a temporary failure of that Physical Volume or an indication that it is failing. The device should be refreshed or replaced. (m)ismatches signifies that the RAID logical volume -has portions of the array that are not coherent or that the array has -recently repaired inconsistencies. An additional "check" after a "repair" -of a RAID logical volume will clear this flag if no additional discrepancies -are found. ("check" and "repair" of a RAID Logical Volume can be done via -the 'lvchange' command.) (w)ritemostly signifies the devices in a RAID 1 -logical volume that have been marked write-mostly. +has portions of the array that are not coherent. Inconsistencies are +detected by initiating a "check" on a RAID logical volume. (The scrubbing +operations, "check" and "repair", can be performed on a RAID logical +volume via the 'lvchange' command.) (w)ritemostly signifies the devices in +a RAID 1 logical volume that have been marked write-mostly. .IP 10 3 s(k)ip activation: this volume is flagged to be skipped during activation. .RE

1 0

← Newer
1
2
3
4
5
6
7
8
...
11
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

lvm2-commits September 2013