Fix statechange-led.sh & unnecessary libdevmapper warning

- Fix autoreplace behaviour on statechange-led.sh script.

ZED sends the following events on an auto-replace:

1. statechange: Disk goes UNAVAIL->ONLINE
2. statechange: Disk goes ONLINE->UNAVAIL
3. vdev_attach: Disk goes ONLINE

Events 1-2 happen when ZED first attempts to do an auto-online.  When that
fails, ZED then tries an auto-replace, generating the vdev_attach event in #3.

In the previous code, statechange-led was only looking at the UNAVAIL->ONLINE
transition to turn off the LED.  It ignored the #2 ONLINE->UNAVAIL transition,
assuming it was just the "old" VDEV going offline.  This is problematic, as
a drive can go from ONLINE->UNAVAIL when it's malfunctioning, and we don't want
to ignore that.

This new patch correctly turns on the fault LED every time a drive becomes
UNAVAIL.  It also monitors vdev_attach events to trigger turning off the LED
when an auto-replaced disk comes online.

- Remove unnecessary libdevmapper warning with --with-config=kernel

This fixes an unnecessary libdevmapper warning when building
--with-config=kernel.  Kernel code does not use libdevmapper, so the warning
is not needed.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes #2375
Closes #5312
Closes #5331

diff --git a/cmd/zed/zed.d/statechange-led.sh b/cmd/zed/zed.d/statechange-led.sh
index af1a14a9af3c02b389dbe43c6fc2751aec285e90..257e4a8a0034d375a6c0dcddeb0d1ab79221ec7d 100755 (executable)
--- a/cmd/zed/zed.d/statechange-led.sh
+++ b/cmd/zed/zed.d/statechange-led.sh
@@ -71,29 +71,12 @@ function led
 function process {
        path="$1"
        fault=$2
-       prev=$3
-       if [ "$fault" == "FAULTED" ] || [ "$fault" == "DEGRADED" ] ; then
-               led "$path" 1
-       elif [ "$fault" == "UNAVAIL" ] && [ "$prev" != "ONLINE" ] ; then
-               # For the most part, UNAVAIL should turn on the LED.  However,
-               # during an autoreplace, we see our new drive go online,
-               # followed by our "old" drive going ONLINE->UNAVAIL. Since the
-               # "old" drive has the same slot information, we want to ignore
-               # the ONLINE->UNAVAIL event.
-               #
-               #       NAME             STATE     READ WRITE CKSUM
-               #       mypool3          DEGRADED     0     0     0
-               #         mirror-0       DEGRADED     0     0     0
-               #           A1           ONLINE       0     0     0
-               #           A2           ONLINE       0   880     0
-               #           replacing-3  UNAVAIL      0     0     0
-               #             old        UNAVAIL      0 2.93K     0  corrupted data
-               #             A3         ONLINE       0     0   156  (resilvering)
+       if [ "$fault" == "FAULTED" ] || [ "$fault" == "DEGRADED" ] || \
+          [ "$fault" == "UNAVAIL" ] ; then
                led "$path" 1
        elif [ "$fault" == "ONLINE" ] ; then
                led "$path" 0
        fi
 }
 
-process "$ZEVENT_VDEV_ENC_SYSFS_PATH" "$ZEVENT_VDEV_STATE_STR" \
-    "$ZEVENT_VDEV_LASTSTATE_STR"
+process "$ZEVENT_VDEV_ENC_SYSFS_PATH" "$ZEVENT_VDEV_STATE_STR"

diff --git a/cmd/zed/zed.d/vdev_attach-led.sh b/cmd/zed/zed.d/vdev_attach-led.sh
new file mode 120000 (symlink)
index 0000000..7d74043
--- /dev/null
+++ b/cmd/zed/zed.d/vdev_attach-led.sh
@@ -0,0 +1 @@
+statechange-led.sh
\ No newline at end of file

diff --git a/configure.ac b/configure.ac
index e9f50f78369275580ad67fe97296ef3367465147..39bd54c17bcd0930b569fc33277f2de316fcea1e 100644 (file)
--- a/configure.ac
+++ b/configure.ac
@@ -304,6 +304,7 @@ AC_CONFIG_FILES([
 
 AC_OUTPUT
 
-AS_IF([test "x$user_libdevmapper" != xyes ], [
-    AC_MSG_WARN([Building without libdevmapper.  Auto-replace, auto-online, and statechange-led.sh may not work correctly with device mapper vdevs.])
+AS_IF([test "x$user_libdevmapper" != xyes && test "$ZFS_CONFIG" != kernel ], [
+    AC_MSG_WARN([Building without libdevmapper.  Auto-replace, auto-online, \
+and statechange-led.sh may not work correctly with device mapper vdevs.])
 ])

diff --git a/module/zfs/zfs_fm.c b/module/zfs/zfs_fm.c
index 0f8ac28c138bb09c32b933bbe09f87c4626e2b56..df37fed2b9729742eb4f83bf76f84e472a40e39f 100644 (file)
--- a/module/zfs/zfs_fm.c
+++ b/module/zfs/zfs_fm.c
@@ -307,6 +307,10 @@ zfs_ereport_start(nvlist_t **ereport_out, nvlist_t **detector_out,
                        fm_payload_set(ereport,
                            FM_EREPORT_PAYLOAD_ZFS_VDEV_FRU,
                            DATA_TYPE_STRING, vd->vdev_fru, NULL);
+               if (vd->vdev_enc_sysfs_path != NULL)
+                       fm_payload_set(ereport,
+                           FM_EREPORT_PAYLOAD_ZFS_VDEV_ENC_SYSFS_PATH,
+                           DATA_TYPE_STRING, vd->vdev_enc_sysfs_path, NULL);
                if (vd->vdev_ashift)
                        fm_payload_set(ereport,
                            FM_EREPORT_PAYLOAD_ZFS_VDEV_ASHIFT,
@@ -928,6 +932,10 @@ zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name,
                if (vd->vdev_fru != NULL)
                        VERIFY0(nvlist_add_string(resource,
                            FM_EREPORT_PAYLOAD_ZFS_VDEV_FRU, vd->vdev_fru));
+               if (vd->vdev_enc_sysfs_path != NULL)
+                       VERIFY0(nvlist_add_string(resource,
+                           FM_EREPORT_PAYLOAD_ZFS_VDEV_ENC_SYSFS_PATH,
+                           vd->vdev_enc_sysfs_path));
                /* also copy any optional payload data */
                if (aux) {
                        nvpair_t *elem = NULL;