rpms/lvm2
6
0
Fork 0
Code Issues Pull Requests Releases Activity

import UBI lvm2-2.03.14-15.el8_10.2

Browse Source
This commit is contained in:
eabdullin 2025-07-15 07:51:37 +00:00
parent 48dfb805f0
commit ea6214f636
3 changed files with 250 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

196
SOURCES/0140-mirror-enhance-error-path-for-pvmove-finish.patch Normal file
View File

@ -0,0 +1,196 @@
From ae515916c05218a49b40afaaf26f9c4c0bb55e53 Mon Sep 17 00:00:00 2001
From: Zdenek Kabelac <zkabelac@redhat.com>
Date: Sat, 8 Feb 2025 21:51:09 +0100
Subject: [PATCH] mirror: enhance error path for pvmove finish
When the pvmove operation is completing, it attempts to deactivate
the temporary mirror and remove its mirror legs. However,
if an external tool holds these volumes open, the operation would
previously abort entirely, leaving the LVM2 metadata in a partially
unusable state that required manual administrative fixes.
To improve this, the code has been enhanced to handle such scenarios
more gracefully. It will now complete the pvmove operation even
if some volumes cannot be deactivated, marking them in the metadata
with an error segment. While the command will report errors,
the metadata will remain in a usable state. The administrator
can then remove the orphaned volumes when they are no longer in use.
(cherry picked from commit ed9468153ec3d9cec8d6fbb9f6b8e09e10427416)
---
lib/metadata/mirror.c | 93 ++++++++++++++++++++++++++++++++++++++++---
tools/pvmove_poll.c | 13 ++++++
2 files changed, 101 insertions(+), 5 deletions(-)
diff --git a/lib/metadata/mirror.c b/lib/metadata/mirror.c
index 46da57948..d5c988372 100644
--- a/lib/metadata/mirror.c
+++ b/lib/metadata/mirror.c
@@ -398,6 +398,59 @@ static int _activate_lv_like_model(struct logical_volume *model,
return 1;
}
+/*
+ * Inherit tags @from_lv to @to_lv, tags maybe needed for activation
+ */
+static int _inherit_lv_tags(const struct logical_volume *from_lv, struct logical_volume *to_lv)
+{
+ struct dm_str_list *sl;
+
+ if (to_lv && !str_list_match_list(&from_lv->tags, &to_lv->tags, NULL))
+ dm_list_iterate_items(sl, &from_lv->tags)
+ if (!str_list_add(from_lv->vg->cmd->mem, &to_lv->tags, sl->str)) {
+ log_error("Aborting. Unable to inherit tag.");
+ return 0;
+ }
+
+ return 1;
+}
+
+/*
+ * Deactivates and removes an 'orphan' temporary @lv,
+ * which is expected to already have an 'error' segment.
+ * Sets @updated_mda (!NULL) to 1 when LV is removed.
+ * Sets @deactivation_failed (!NULL) to 1 when deactivation fails.
+ *
+ * Note: An external tool might still have the volume open, preventing
+ * immediate deactivation. If deactivation fails (even after retrying),
+ * it is safer to proceed with the command and leave the LV visible.
+ * This allows the user to manually remove it when it is no longer in use.
+ *
+ * Any errors will be detected later in the process, as there will be
+ * more visible LVs than expected.
+ */
+static int _deactivate_and_remove_lv(struct logical_volume *lv,
+ int *updated_mda,
+ int *deactivation_failed)
+{
+ if (lv) {
+ /* FIXME: convert to use lv_active_change() */
+ if (!deactivate_lv(lv->vg->cmd, lv)) {
+ /* Note: still returns success here and fails later */
+ log_warn("WARNING: Can't deactivate temporary volume %s.",
+ display_lvname(lv));
+ if (deactivation_failed)
+ *deactivation_failed = 1;
+ } else if (!lv_remove(lv)) {
+ /* Can't continue with internal metadata problems */
+ return_0;
+ } else if (updated_mda)
+ *updated_mda = 1;
+ }
+
+ return 1;
+}
+
/*
* Delete independent/orphan LV, it must acquire lock.
*/
@@ -803,7 +856,6 @@ static int _remove_mirror_images(struct logical_volume *lv,
struct lv_list *lvl;
struct dm_list tmp_orphan_lvs;
uint32_t orig_removed = num_removed;
- int reactivate;
if (removed)
*removed = 0;
@@ -967,6 +1019,19 @@ static int _remove_mirror_images(struct logical_volume *lv,
return_0;
}
+ if (!collapse) {
+ dm_list_iterate_items(lvl, &tmp_orphan_lvs) {
+ if (!_inherit_lv_tags(lv, lvl->lv))
+ return_0;
+ if (!replace_lv_with_error_segment(lvl->lv))
+ return_0;
+ }
+ }
+
+ if (!_inherit_lv_tags(lv, temp_layer_lv) ||
+ !_inherit_lv_tags(lv, detached_log_lv))
+ return_0;
+
/*
* To successfully remove these unwanted LVs we need to
* remove the LVs from the mirror set, commit that metadata
@@ -976,17 +1041,35 @@ static int _remove_mirror_images(struct logical_volume *lv,
return_0;
/* Save or delete the 'orphan' LVs */
- reactivate = lv_is_active(lv_lock_holder(lv));
+ if (lv_is_active(lv_lock_holder(lv))) {
+ if (!collapse) {
+ dm_list_iterate_items(lvl, &tmp_orphan_lvs)
+ if (!_activate_lv_like_model(lv, lvl->lv))
+ return_0;
+ }
+
+ if (temp_layer_lv &&
+ !_activate_lv_like_model(lv, temp_layer_lv))
+ return_0;
+
+ if (detached_log_lv &&
+ !_activate_lv_like_model(lv, detached_log_lv))
+ return_0;
+
+ if (!sync_local_dev_names(lv->vg->cmd))
+ stack;
+ }
+
if (!collapse) {
dm_list_iterate_items(lvl, &tmp_orphan_lvs)
- if (!_delete_lv(lv, lvl->lv, reactivate))
+ if (!_deactivate_and_remove_lv(lvl->lv, NULL, NULL))
return_0;
}
- if (temp_layer_lv && !_delete_lv(lv, temp_layer_lv, reactivate))
+ if (!_deactivate_and_remove_lv(temp_layer_lv, NULL, NULL))
return_0;
- if (detached_log_lv && !_delete_lv(lv, detached_log_lv, reactivate))
+ if (!_deactivate_and_remove_lv(detached_log_lv, NULL, NULL))
return_0;
/* Mirror with only 1 area is 'in sync'. */
diff --git a/tools/pvmove_poll.c b/tools/pvmove_poll.c
index 751313cd7..8b97905c0 100644
--- a/tools/pvmove_poll.c
+++ b/tools/pvmove_poll.c
@@ -87,6 +87,8 @@ int pvmove_update_metadata(struct cmd_context *cmd, struct volume_group *vg,
int pvmove_finish(struct cmd_context *cmd, struct volume_group *vg,
struct logical_volume *lv_mirr, struct dm_list *lvs_changed)
{
+ uint32_t visible = vg_visible_lvs(lv_mirr->vg);
+
if (!dm_list_empty(lvs_changed) &&
(!_detach_pvmove_mirror(cmd, lv_mirr) ||
!replace_lv_with_error_segment(lv_mirr))) {
@@ -94,6 +96,8 @@ int pvmove_finish(struct cmd_context *cmd, struct volume_group *vg,
return 0;
}
+ lv_set_visible(lv_mirr);
+
if (!lv_update_and_reload(lv_mirr))
return_0;
@@ -120,5 +124,14 @@ int pvmove_finish(struct cmd_context *cmd, struct volume_group *vg,
return 0;
}
+ /* Allows the pvmove operation to complete even if 'orphaned' temporary volumes
+ * cannot be deactivated due to being held open by another process.
+ * The user can manually remove these volumes later when they are no longer in use. */
+ if (visible < vg_visible_lvs(lv_mirr->vg)) {
+ log_error("ABORTING: Failed to remove temporary logical volume(s).");
+ log_print_unless_silent("Please remove orphan temporary logical volume(s) when possible.");
+ return 0;
+ }
+
return 1;
}
--
2.48.1

42
SOURCES/0141-lvmlockd-vgchange-systemid-doen-t-need-global-lock.patch Normal file
View File

@ -0,0 +1,42 @@
From 97aa5f67547f150b5f8f511249b200a41561a3c9 Mon Sep 17 00:00:00 2001
From: David Teigland <teigland@redhat.com>
Date: Thu, 27 Jun 2024 12:46:15 -0500
Subject: [PATCH 1/2] lvmlockd: vgchange systemid doen't need global lock
The comment explained that the ex global lock was just
used to trigger global cache invalidation, which is no
longer needed. This extra locking can cause problems
with LVM-activate when local and shared VGs are mixed
(and the incorrect exit code for errors was causing
problems.)
(cherry picked from commit 680f7bd676526771b7a26e35105af3bccb95e4a1)
---
tools/vgchange.c | 11 -----------
1 file changed, 11 deletions(-)
diff --git a/tools/vgchange.c b/tools/vgchange.c
index e4b57dbd2..b0e2b2d2a 100644
--- a/tools/vgchange.c
+++ b/tools/vgchange.c
@@ -1231,17 +1231,6 @@ int vgchange_systemid_cmd(struct cmd_context *cmd, int argc, char **argv)
struct processing_handle *handle;
int ret;
- /*
- * This is a special case where taking the global lock is
- * not needed to protect global state, because the change is
- * only to an existing VG. But, taking the global lock ex is
- * helpful in this case to trigger a global cache validation
- * on other hosts, to cause them to see the new system_id or
- * lock_type.
- */
- if (!lockd_global(cmd, "ex"))
- return 0;
-
if (!(handle = init_processing_handle(cmd, NULL))) {
log_error("Failed to initialize processing handle.");
return ECMD_FAILED;
--
2.49.0

14
SPECS/lvm2.spec
View File

@ -55,7 +55,7 @@
%global commit 4dc5d4ac7e7a9457ccc46ff04796b347e58bf4da %global commit 4dc5d4ac7e7a9457ccc46ff04796b347e58bf4da
%global shortcommit %(c=%{commit}; echo ${c:0:7}) %global shortcommit %(c=%{commit}; echo ${c:0:7})
%endif %endif
#%%global rel_suffix .bz2233901_1 %global rel_suffix .2
# Do not reset Release to 1 unless both lvm2 and device-mapper # Do not reset Release to 1 unless both lvm2 and device-mapper
# versions are increased together. # versions are increased together.
@ -237,6 +237,10 @@ Patch137: 0136-configure-autoreconf.patch
Patch138: 0137-dmeventd-implement-exit_on-file-check.patch Patch138: 0137-dmeventd-implement-exit_on-file-check.patch
Patch139: 0138-debug-correct-level.patch Patch139: 0138-debug-correct-level.patch
Patch140: 0139-tests-check-exit_on-works.patch Patch140: 0139-tests-check-exit_on-works.patch
# RHEL-62764:
Patch141: 0140-mirror-enhance-error-path-for-pvmove-finish.patch
# RHEL-66486:
Patch142: 0141-lvmlockd-vgchange-systemid-doen-t-need-global-lock.patch
BuildRequires: gcc BuildRequires: gcc
%if %{enable_testsuite} %if %{enable_testsuite}
@ -902,7 +906,7 @@ the device-mapper event library.
%package testsuite %package testsuite
Summary: LVM2 Testsuite Summary: LVM2 Testsuite
# Most of the code is GPLv2, the harness in test/lib/{brick-shelltest.h,runner.cpp} is BSD, and C files in test/api are LGPLv2... # Most of the code is GPLv2, the harness in test/lib/{brick-shelltest.h,runner.cpp} is BSD, and C files in test/api are LGPLv2...
License: LGPLv2 and GPLv2 and BSD-2-Clause License: GPL-2.0-only AND LGPL-2.1-only AND BSD-2-Clause
%description testsuite %description testsuite
An extensive functional testsuite for LVM2. An extensive functional testsuite for LVM2.
@ -915,6 +919,12 @@ An extensive functional testsuite for LVM2.
%endif %endif
%changelog %changelog
* Mon Jun 02 2025 Marian Csontos <mcsontos@redhat.com> - 2.03.14-15.el8_10.2
- Fix vgchange exit code when lvmlockd is not running and no change is done.
* Wed Apr 30 2025 Marian Csontos <mcsontos@redhat.com> - 2.03.14-15.el8_10.1
- Fix incomplete pvmove write bad metadata requiring manual intervention.
* Wed Jan 22 2025 Marian Csontos <mcsontos@redhat.com> - 2.03.14-15 * Wed Jan 22 2025 Marian Csontos <mcsontos@redhat.com> - 2.03.14-15
- Fix dmeventd blocking on shutdown. - Fix dmeventd blocking on shutdown.
- Force exit dmeventd when /run/nologin is present. - Force exit dmeventd when /run/nologin is present.