c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2026-05-22 18:40:43 +03:00
Code Issues Packages Projects Releases Wiki Activity

zed: Ensure spare activation after kernel-initiated device removal

Browse Source 
In addition to hotplug events, the kernel may also mark a failing vdev
as REMOVED. This was observed in a customer report and reproduced by
forcing the NVMe host driver to disable the device after a failed reset
due to command timeout. In such cases, the spare was not activated
because the device had already transitioned to a REMOVED state before
zed processed the event.
To address this, explicitly attempt hot spare activation when the
kernel marks a device as REMOVED.

Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Ameer Hamza <ahamza@ixsystems.com>
Closes #17187
This commit is contained in:
Ameer Hamza
2025-03-29 00:48:38 +05:00
committed by GitHub
parent dd2a46b5e6
commit 30cc2331f4
5 changed files with 46 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
module/zfs/spa.c
+12 -7
View File
@@ -8921,7 +8921,7 @@ spa_scan_range(spa_t *spa, pool_scan_func_t func, uint64_t txgstart,
*/
static void
spa_async_remove(spa_t *spa, vdev_t *vd)
spa_async_remove(spa_t *spa, vdev_t *vd, boolean_t by_kernel)
{
if (vd->vdev_remove_wanted) {
vd->vdev_remove_wanted = B_FALSE;
@@ -8941,11 +8941,11 @@ spa_async_remove(spa_t *spa, vdev_t *vd)
vdev_state_dirty(vd->vdev_top);
/* Tell userspace that the vdev is gone. */
zfs_post_remove(spa, vd);
zfs_post_remove(spa, vd, by_kernel);
}
for (int c = 0; c < vd->vdev_children; c++)
spa_async_remove(spa, vd->vdev_child[c]);
spa_async_remove(spa, vd->vdev_child[c], by_kernel);
}
static void
@@ -9039,13 +9039,18 @@ spa_async_thread(void *arg)
/*
* See if any devices need to be marked REMOVED.
*/
if (tasks & SPA_ASYNC_REMOVE) {
if (tasks & (SPA_ASYNC_REMOVE | SPA_ASYNC_REMOVE_BY_USER)) {
boolean_t by_kernel = B_TRUE;
if (tasks & SPA_ASYNC_REMOVE_BY_USER)
by_kernel = B_FALSE;
spa_vdev_state_enter(spa, SCL_NONE);
spa_async_remove(spa, spa->spa_root_vdev);
spa_async_remove(spa, spa->spa_root_vdev, by_kernel);
for (int i = 0; i < spa->spa_l2cache.sav_count; i++)
spa_async_remove(spa, spa->spa_l2cache.sav_vdevs[i]);
spa_async_remove(spa, spa->spa_l2cache.sav_vdevs[i],
by_kernel);
for (int i = 0; i < spa->spa_spares.sav_count; i++)
spa_async_remove(spa, spa->spa_spares.sav_vdevs[i]);
spa_async_remove(spa, spa->spa_spares.sav_vdevs[i],
by_kernel);
(void) spa_vdev_state_exit(spa, NULL, 0);
}
module/zfs/vdev.c
+1 -1
View File
@@ -4271,7 +4271,7 @@ vdev_remove_wanted(spa_t *spa, uint64_t guid)
return (spa_vdev_state_exit(spa, NULL, SET_ERROR(EEXIST)));
vd->vdev_remove_wanted = B_TRUE;
spa_async_request(spa, SPA_ASYNC_REMOVE);
spa_async_request(spa, SPA_ASYNC_REMOVE_BY_USER);
return (spa_vdev_state_exit(spa, vd, 0));
}
module/zfs/zfs_fm.c
+16 -2
View File
@@ -1433,9 +1433,23 @@ zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name,
* removal.
*/
void
zfs_post_remove(spa_t *spa, vdev_t *vd)
zfs_post_remove(spa_t *spa, vdev_t *vd, boolean_t by_kernel)
{
zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_REMOVED, NULL);
nvlist_t *aux = NULL;
if (by_kernel) {
/*
* Add optional supplemental keys to payload
*/
aux = fm_nvlist_create(NULL);
if (aux)
fnvlist_add_boolean(aux, "by_kernel");
}
zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_REMOVED, aux);
if (by_kernel && aux)
fm_nvlist_destroy(aux, FM_NVA_FREE);
}
/*