rt/snapraid
1
0
Fork 0
mirror of https://github.com/amadvance/snapraid.git synced 2026-10-26 10:51:54 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

Get SMART attributes also in probe, but only if the disk is already spinning

Browse source
This commit is contained in:
Andrea Mazzoleni 2025-12-25 21:03:46 +01:00
parent 986d7464b4
commit fca525e5c6
4 changed files with 44 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

40
cmdline/unix.c
View file

@ -1533,7 +1533,7 @@ static void devattr(dev_t device, uint64_t* info, char* serial, char* vendor, ch
* Get POWER state.
*/
#if HAVE_LINUX_DEVICE
static int devprobe(dev_t device, const char* name, const char* smartctl, int* power, uint64_t* info, char* serial, char* vendor, char* model)
static int devprobe(dev_t device, const char* name, const char* smartctl, int* power, uint64_t* smart, uint64_t* info, char* serial, char* vendor, char* model)
{
char cmd[PATH_MAX + 64];
char file[PATH_MAX];
@ -1560,9 +1560,9 @@ static int devprobe(dev_t device, const char* name, const char* smartctl, int* p
if (smartctl[0]) {
char option[PATH_MAX];
snprintf(option, sizeof(option), smartctl, file);
snprintf(cmd, sizeof(cmd), "%s -n standby,3 -i %s", x, option);
snprintf(cmd, sizeof(cmd), "%s -n standby,3 -a %s", x, option);
} else {
snprintf(cmd, sizeof(cmd), "%s -n standby,3 -i %s", x, file);
snprintf(cmd, sizeof(cmd), "%s -n standby,3 -a %s", x, file);
}
log_tag("smartctl:%s:%s:run: %s\n", file, name, cmd);
@ -1576,7 +1576,7 @@ static int devprobe(dev_t device, const char* name, const char* smartctl, int* p
/* LCOV_EXCL_STOP */
}
if (smartctl_attribute(f, file, name, 0, info, serial, vendor, model) != 0) {
if (smartctl_attribute(f, file, name, smart, info, serial, vendor, model) != 0) {
/* LCOV_EXCL_START */
pclose(f);
log_tag("device:%s:%s:shell\n", file, name);
@ -1603,18 +1603,15 @@ static int devprobe(dev_t device, const char* name, const char* smartctl, int* p
/* LCOV_EXCL_STOP */
}
if (WEXITSTATUS(ret) == 0) {
log_tag("attr:%s:%s:power:active\n", file, name);
*power = POWER_ACTIVE;
} else if (WEXITSTATUS(ret) == 3) {
if (WEXITSTATUS(ret) == 3) {
log_tag("attr:%s:%s:power:standby\n", file, name);
*power = POWER_STANDBY;
} else {
/* LCOV_EXCL_START */
log_tag("device:%s:%s:exit:%d\n", file, name, WEXITSTATUS(ret));
log_fatal("Failed to run '%s' with return code %xh.\n", cmd, WEXITSTATUS(ret));
return -1;
/* LCOV_EXCL_STOP */
log_tag("attr:%s:%s:power:active\n", file, name);
*power = POWER_ACTIVE;
/* store the return smartctl return value */
smart[SMART_FLAGS] = WEXITSTATUS(ret);
}
return 0;
@ -1716,7 +1713,7 @@ static int devdownifup(dev_t device, const char* name, const char* smartctl, int
{
*power = POWER_UNKNOWN;
if (devprobe(device, name, smartctl, power, 0, 0, 0, 0) != 0)
if (devprobe(device, name, smartctl, power, 0, 0, 0, 0, 0) != 0)
return -1;
if (*power == POWER_ACTIVE)
@ -1888,6 +1885,19 @@ static void* thread_smart(void* arg)
/* LCOV_EXCL_STOP */
}
/*
* Retrieve some attributes directly from the system.
*
* smartctl intentionally skips queries on devices in standby mode
* to prevent accidentally spinning them up.
*/
devattr(devinfo->device, devinfo->info, devinfo->serial, devinfo->family, devinfo->model);
/*
* Retrieve access stat for the device
*/
devstat(devinfo->device, &devinfo->access_stat);
return 0;
#else
(void)arg;
@ -1903,7 +1913,7 @@ static void* thread_probe(void* arg)
#if HAVE_LINUX_DEVICE
devinfo_t* devinfo = arg;
if (devprobe(devinfo->device, devinfo->name, devinfo->smartctl, &devinfo->power, devinfo->info, devinfo->serial, devinfo->family, devinfo->model) != 0) {
if (devprobe(devinfo->device, devinfo->name, devinfo->smartctl, &devinfo->power, devinfo->smart, devinfo->info, devinfo->serial, devinfo->family, devinfo->model) != 0) {
/* LCOV_EXCL_START */
return (void*)-1;
/* LCOV_EXCL_STOP */

8
doc/snapraid_log.1
View file

@ -1435,8 +1435,12 @@ Logs the final exit status indicating that the current hash
algorithm is already the \`best hash\` for the platform, and thus
a rehash is not required. This causes the program to exit with
a fatal error.
.SH COMMAND SMART TAGS
This section describes the tags output with the \`smart\` command.
.SH COMMAND SMART AND PROBE TAGS
This section describes the tags output by the \`smart\` and \`probe\`
commands. The main difference between the two is that \`probe\` doesn\'t
spin up the disk, so it won\'t be able to get all attributes if the
disk is spun down. The \`smart\` command, instead, will always get all
attributes, but with the risk of spinning up all disks.
.PP
Note that a \`<disk_name>\` may be associated with more \`<device_file>\`,
in case the logical disk uses more physical disks. In such a case,

8
doc/snapraid_log.d
View file

@ -917,8 +917,12 @@ Command Rehash Tags
a rehash is not required. This causes the program to exit with
a fatal error.
Command Smart Tags
This section describes the tags output with the `smart` command.
Command Smart And Probe Tags
This section describes the tags output by the `smart` and `probe`
commands. The main difference between the two is that `probe` doesn't
spin up the disk, so it won't be able to get all attributes if the
disk is spun down. The `smart` command, instead, will always get all
attributes, but with the risk of spinning up all disks.
Note that a `<disk_name>` may be associated with more `<device_file>`,
in case the logical disk uses more physical disks. In such a case,

10
doc/snapraid_log.txt
View file

@ -957,10 +957,14 @@ These tags provide a final summary of the rehash scheduling process.
a fatal error.
14 COMMAND SMART TAGS
=====================
14 COMMAND SMART AND PROBE TAGS
===============================
This section describes the tags output with the `smart` command.
This section describes the tags output by the `smart` and `probe`
commands. The main difference between the two is that `probe` doesn't
spin up the disk, so it won't be able to get all attributes if the
disk is spun down. The `smart` command, instead, will always get all
attributes, but with the risk of spinning up all disks.
Note that a `<disk_name>` may be associated with more `<device_file>`,
in case the logical disk uses more physical disks. In such a case,