#include <limits.h>
#include <syslog.h>
-/* The largest number of disks current arrays can manage is 384
- * This really should be dynamically, but that will have to wait
- * At least it isn't MD_SB_DISKS.
- */
-#define MaxDisks 384
struct state {
char *devname;
int devnum; /* to sync with mdstat info */
char *spare_group;
int active, working, failed, spare, raid;
int expected_spares;
- int devstate[MaxDisks];
- dev_t devid[MaxDisks];
+ int devstate[MAX_DISKS];
+ dev_t devid[MAX_DISKS];
int percent;
int parent_dev; /* For subarray, devnum of parent.
* For others, NoMdDev
static void alert(char *event, char *dev, char *disc, struct alert_info *info);
static int check_array(struct state *st, struct mdstat_ent *mdstat,
int test, struct alert_info *info,
- int increments);
+ int increments, char *prefer);
static int add_new_arrays(struct mdstat_ent *mdstat, struct state **statelist,
int test, struct alert_info *info);
static void try_spare_migration(struct state *statelist, struct alert_info *info);
char *mailaddr, char *alert_cmd,
int period, int daemonise, int scan, int oneshot,
int dosyslog, int test, char *pidfile, int increments,
- int share)
+ int share, char *prefer)
{
/*
* Every few seconds, scan every md device looking for changes
*/
struct state *statelist = NULL;
+ struct state *st2;
int finished = 0;
struct mdstat_ent *mdstat = NULL;
char *mailfrom = NULL;
if (!mailaddr) {
mailaddr = conf_get_mailaddr();
if (mailaddr && ! scan)
- fprintf(stderr, Name ": Monitor using email address \"%s\" from config file\n",
+ pr_err("Monitor using email address \"%s\" from config file\n",
mailaddr);
}
mailfrom = conf_get_mailfrom();
if (!alert_cmd) {
alert_cmd = conf_get_program();
if (alert_cmd && ! scan)
- fprintf(stderr, Name ": Monitor using program \"%s\" from config file\n",
+ pr_err("Monitor using program \"%s\" from config file\n",
alert_cmd);
}
- if (scan && !mailaddr && !alert_cmd) {
- fprintf(stderr, Name ": No mail address or alert command - not monitoring.\n");
+ if (scan && !mailaddr && !alert_cmd && !dosyslog) {
+ pr_err("No mail address or alert command - not monitoring.\n");
return 1;
}
info.alert_cmd = alert_cmd;
continue;
if (strcasecmp(mdlist->devname, "<ignore>") == 0)
continue;
- st = calloc(1, sizeof *st);
- if (st == NULL)
- continue;
+ st = xcalloc(1, sizeof *st);
if (mdlist->devname[0] == '/')
- st->devname = strdup(mdlist->devname);
+ st->devname = xstrdup(mdlist->devname);
else {
- st->devname = malloc(8+strlen(mdlist->devname)+1);
+ st->devname = xmalloc(8+strlen(mdlist->devname)+1);
strcpy(strcpy(st->devname, "/dev/md/"),
mdlist->devname);
}
st->next = statelist;
st->devnum = INT_MAX;
- st->percent = -2;
+ st->percent = RESYNC_UNKNOWN;
st->expected_spares = mdlist->spare_disks;
if (mdlist->spare_group)
- st->spare_group = strdup(mdlist->spare_group);
+ st->spare_group = xstrdup(mdlist->spare_group);
statelist = st;
}
} else {
struct mddev_dev *dv;
for (dv=devlist ; dv; dv=dv->next) {
struct mddev_ident *mdlist = conf_get_ident(dv->devname);
- struct state *st = calloc(1, sizeof *st);
- if (st == NULL)
- continue;
- st->devname = strdup(dv->devname);
+ struct state *st = xcalloc(1, sizeof *st);
+ st->devname = xstrdup(dv->devname);
st->next = statelist;
st->devnum = INT_MAX;
- st->percent = -2;
+ st->percent = RESYNC_UNKNOWN;
st->expected_spares = -1;
if (mdlist) {
st->expected_spares = mdlist->spare_disks;
if (mdlist->spare_group)
- st->spare_group = strdup(mdlist->spare_group);
+ st->spare_group = xstrdup(mdlist->spare_group);
}
statelist = st;
}
mdstat = mdstat_read(oneshot?0:1, 0);
for (st=statelist; st; st=st->next)
- if (check_array(st, mdstat, test, &info, increments))
+ if (check_array(st, mdstat, test, &info,
+ increments, prefer))
anydegraded = 1;
/* now check if there are any new devices found in mdstat */
}
test = 0;
}
+ for (st2 = statelist; st2; st2 = statelist) {
+ statelist = st2->next;
+ free(st2);
+ }
+
if (pidfile)
unlink(pidfile);
return 0;
int pid, rv;
FILE *fp;
char dir[20];
+ char path[100];
struct stat buf;
- fp = fopen("/var/run/mdadm/autorebuild.pid", "r");
+ sprintf(path, "%s/autorebuild.pid", MDMON_DIR);
+ fp = fopen(path, "r");
if (fp) {
if (fscanf(fp, "%d", &pid) != 1)
pid = -1;
rv = stat(dir, &buf);
if (rv != -1) {
if (scan) {
- fprintf(stderr, Name ": Only one "
+ pr_err("Only one "
"autorebuild process allowed"
" in scan mode, aborting\n");
fclose(fp);
return 1;
} else {
- fprintf(stderr, Name ": Warning: One"
+ pr_err("Warning: One"
" autorebuild process already"
" running.\n");
}
fclose(fp);
}
if (scan) {
- if (mkdir("/var/run/mdadm", S_IRWXU) < 0 &&
+ if (mkdir(MDMON_DIR, S_IRWXU) < 0 &&
errno != EEXIST) {
- fprintf(stderr, Name ": Can't create "
+ pr_err("Can't create "
"autorebuild.pid file\n");
} else {
- fp = fopen("/var/run/mdadm/autorebuild.pid", "w");
+ fp = fopen(path, "w");
if (!fp)
- fprintf(stderr, Name ": Cannot create"
+ pr_err("Cannot create"
" autorebuild.pid"
"file\n");
else {
{
int priority;
- if (!info->alert_cmd && !info->mailaddr) {
+ if (!info->alert_cmd && !info->mailaddr && !info->dosyslog) {
time_t now = time(0);
printf("%1.15s: %s on %s %s\n", ctime(&now)+4, event, dev, disc?disc:"unknown device");
static int check_array(struct state *st, struct mdstat_ent *mdstat,
int test, struct alert_info *ainfo,
- int increments)
+ int increments, char *prefer)
{
/* Update the state 'st' to reflect any changes shown in mdstat,
* or found by directly examining the array, and return
* '1' if the array is degraded, or '0' if it is optimal (or dead).
*/
- struct { int state, major, minor; } info[MaxDisks];
+ struct { int state, major, minor; } info[MAX_DISKS];
mdu_array_info_t array;
struct mdstat_ent *mse = NULL, *mse2;
char *dev = st->devname;
int fd;
int i;
+ int remaining_disks;
+ int last_disk;
+ int new_array = 0;
if (test)
alert("TestMessage", dev, NULL, ainfo);
* just make sure it is always different. */
array.utime = st->utime + 1;;
+ if (st->err) {
+ /* New array appeared where previously had and error */
+ st->err = 0;
+ st->percent = RESYNC_NONE;
+ new_array = 1;
+ alert("NewArray", st->devname, NULL, ainfo);
+ }
+
if (st->utime == array.utime &&
st->failed == array.failed_disks &&
st->working == array.working_disks &&
mse->percent == st->percent
))) {
close(fd);
- st->err = 0;
if ((st->active < st->raid) && st->spare == 0)
return 1;
else
st->expected_spares > 0 &&
array.spare_disks < st->expected_spares)
alert("SparesMissing", dev, NULL, ainfo);
- if (st->percent == -1 &&
+ if (st->percent < 0 && st->percent != RESYNC_UNKNOWN &&
mse->percent >= 0)
alert("RebuildStarted", dev, NULL, ainfo);
if (st->percent >= 0 &&
alert(percentalert, dev, NULL, ainfo);
}
- if (mse->percent == -1 &&
+ if (mse->percent == RESYNC_NONE &&
st->percent >= 0) {
/* Rebuild/sync/whatever just finished.
* If there is a number in /mismatch_cnt,
struct mdinfo *sra =
sysfs_read(-1, st->devnum, GET_MISMATCH);
if (sra && sra->mismatch_cnt > 0) {
- char cnt[40];
- sprintf(cnt, " mismatches found: %d", sra->mismatch_cnt);
+ char cnt[80];
+ snprintf(cnt, sizeof(cnt),
+ " mismatches found: %d (on raid level %d)",
+ sra->mismatch_cnt, array.level);
alert("RebuildFinished", dev, cnt, ainfo);
} else
alert("RebuildFinished", dev, NULL, ainfo);
}
st->percent = mse->percent;
- for (i=0; i<MaxDisks && i <= array.raid_disks + array.nr_disks;
+ remaining_disks = array.nr_disks;
+ for (i=0; i<MAX_DISKS && remaining_disks > 0;
i++) {
mdu_disk_info_t disc;
disc.number = i;
info[i].state = disc.state;
info[i].major = disc.major;
info[i].minor = disc.minor;
+ if (disc.major || disc.minor)
+ remaining_disks --;
} else
info[i].major = info[i].minor = 0;
}
+ last_disk = i;
- if (strncmp(mse->metadata_version, "external:", 9) == 0 &&
+ if (mse->metadata_version &&
+ strncmp(mse->metadata_version, "external:", 9) == 0 &&
is_subarray(mse->metadata_version+9))
st->parent_dev =
devname2devnum(mse->metadata_version+10);
close(fd);
- for (i=0; i<MaxDisks; i++) {
+ for (i=0; i<MAX_DISKS; i++) {
mdu_disk_info_t disc = {0,0,0,0,0};
int newstate=0;
int change;
char *dv = NULL;
disc.number = i;
- if (i > array.raid_disks + array.nr_disks) {
- newstate = 0;
- disc.major = disc.minor = 0;
- } else if (info[i].major || info[i].minor) {
+ if (i < last_disk &&
+ (info[i].major || info[i].minor)) {
newstate = info[i].state;
- dv = map_dev(info[i].major, info[i].minor, 1);
+ dv = map_dev_preferred(
+ info[i].major, info[i].minor, 1,
+ prefer);
disc.state = newstate;
disc.major = info[i].major;
disc.minor = info[i].minor;
- } else if (mse && mse->pattern && i < (int)strlen(mse->pattern)) {
- switch(mse->pattern[i]) {
- case 'U': newstate = 6 /* ACTIVE/SYNC */; break;
- case '_': newstate = 0; break;
- }
- disc.major = disc.minor = 0;
- }
+ } else
+ newstate = (1 << MD_DISK_REMOVED);
+
if (dv == NULL && st->devid[i])
- dv = map_dev(major(st->devid[i]),
- minor(st->devid[i]), 1);
+ dv = map_dev_preferred(
+ major(st->devid[i]),
+ minor(st->devid[i]), 1, prefer);
change = newstate ^ st->devstate[i];
- if (st->utime && change && !st->err) {
- if (i < array.raid_disks &&
- (((newstate&change)&(1<<MD_DISK_FAULTY)) ||
- ((st->devstate[i]&change)&(1<<MD_DISK_ACTIVE)) ||
- ((st->devstate[i]&change)&(1<<MD_DISK_SYNC)))
- )
+ if (st->utime && change && !st->err && !new_array) {
+ if ((st->devstate[i]&change)&(1<<MD_DISK_SYNC))
alert("Fail", dev, dv, ainfo);
- else if (i >= array.raid_disks &&
+ else if ((newstate & (1<<MD_DISK_FAULTY)) &&
(disc.major || disc.minor) &&
- st->devid[i] == makedev(disc.major, disc.minor) &&
- ((newstate&change)&(1<<MD_DISK_FAULTY))
- )
+ st->devid[i] == makedev(disc.major, disc.minor))
alert("FailSpare", dev, dv, ainfo);
- else if (i < array.raid_disks &&
- ! (newstate & (1<<MD_DISK_REMOVED)) &&
- (((st->devstate[i]&change)&(1<<MD_DISK_FAULTY)) ||
- ((newstate&change)&(1<<MD_DISK_ACTIVE)) ||
- ((newstate&change)&(1<<MD_DISK_SYNC)))
- )
+ else if ((newstate&change)&(1<<MD_DISK_SYNC))
alert("SpareActive", dev, dv, ainfo);
}
st->devstate[i] = newstate;
(strcmp(mse->level, "raid0") != 0 &&
strcmp(mse->level, "linear") != 0))
) {
- struct state *st = calloc(1, sizeof *st);
+ struct state *st = xcalloc(1, sizeof *st);
mdu_array_info_t array;
int fd;
- if (st == NULL)
- continue;
- st->devname = strdup(get_md_name(mse->devnum));
+ st->devname = xstrdup(get_md_name(mse->devnum));
if ((fd = open(st->devname, O_RDONLY)) < 0 ||
ioctl(fd, GET_ARRAY_INFO, &array)< 0) {
/* no such array */
st->next = *statelist;
st->err = 1;
st->devnum = mse->devnum;
- st->percent = -2;
+ st->percent = RESYNC_UNKNOWN;
st->expected_spares = -1;
- if (strncmp(mse->metadata_version, "external:", 9) == 0 &&
+ if (mse->metadata_version &&
+ strncmp(mse->metadata_version, "external:", 9) == 0 &&
is_subarray(mse->metadata_version+9))
st->parent_dev =
devname2devnum(mse->metadata_version+10);
*statelist = st;
if (test)
alert("TestMessage", st->devname, NULL, info);
- alert("NewArray", st->devname, NULL, info);
new_found = 1;
}
return new_found;
int d;
dev_t dev = 0;
- for (d = from->raid; !dev && d < MaxDisks; d++) {
+ for (d = from->raid; !dev && d < MAX_DISKS; d++) {
if (from->devid[d] > 0 &&
from->devstate[d] == 0) {
struct dev_policy *pol;
if (devid > 0)
continue;
}
- for (d = 0; d < MaxDisks; d++)
+ for (d = 0; d < MAX_DISKS; d++)
if (to->devid[d])
domainlist_add_dev(&domlist,
to->devid[d],
int rv = 1;
if (stat(dev, &stb) != 0) {
- fprintf(stderr, Name ": Cannot find %s: %s\n", dev,
+ pr_err("Cannot find %s: %s\n", dev,
strerror(errno));
return 2;
}
if (is_subarray(&e->metadata_version[9]))
ping_monitor(&e->metadata_version[9]);
else
- ping_monitor(devnum2devname(devnum));
+ ping_monitor_by_id(devnum);
}
free_mdstat(ms);
return rv;
mdstat_wait(5);
}
}
+
+#ifndef MDASSEMBLE
+
+static char *clean_states[] = {
+ "clear", "inactive", "readonly", "read-auto", "clean", NULL };
+
+int WaitClean(char *dev, int sock, int verbose)
+{
+ int fd;
+ struct mdinfo *mdi;
+ int rv = 1;
+ int devnum;
+
+ fd = open(dev, O_RDONLY);
+ if (fd < 0) {
+ if (verbose)
+ pr_err("Couldn't open %s: %s\n", dev, strerror(errno));
+ return 1;
+ }
+
+ devnum = fd2devnum(fd);
+ mdi = sysfs_read(fd, devnum, GET_VERSION|GET_LEVEL|GET_SAFEMODE);
+ if (!mdi) {
+ if (verbose)
+ pr_err("Failed to read sysfs attributes for "
+ "%s\n", dev);
+ close(fd);
+ return 0;
+ }
+
+ switch(mdi->array.level) {
+ case LEVEL_LINEAR:
+ case LEVEL_MULTIPATH:
+ case 0:
+ /* safemode delay is irrelevant for these levels */
+ rv = 0;
+ }
+
+ /* for internal metadata the kernel handles the final clean
+ * transition, containers can never be dirty
+ */
+ if (!is_subarray(mdi->text_version))
+ rv = 0;
+
+ /* safemode disabled ? */
+ if (mdi->safe_mode_delay == 0)
+ rv = 0;
+
+ if (rv) {
+ int state_fd = sysfs_open(fd2devnum(fd), NULL, "array_state");
+ char buf[20];
+ fd_set fds;
+ struct timeval tm;
+
+ /* minimize the safe_mode_delay and prepare to wait up to 5s
+ * for writes to quiesce
+ */
+ sysfs_set_safemode(mdi, 1);
+ tm.tv_sec = 5;
+ tm.tv_usec = 0;
+
+ FD_ZERO(&fds);
+
+ /* wait for array_state to be clean */
+ while (1) {
+ rv = read(state_fd, buf, sizeof(buf));
+ if (rv < 0)
+ break;
+ if (sysfs_match_word(buf, clean_states) <= 4)
+ break;
+ FD_SET(state_fd, &fds);
+ rv = select(state_fd + 1, NULL, NULL, &fds, &tm);
+ if (rv < 0 && errno != EINTR)
+ break;
+ lseek(state_fd, 0, SEEK_SET);
+ }
+ if (rv < 0)
+ rv = 1;
+ else if (fping_monitor(sock) == 0 ||
+ ping_monitor(mdi->text_version) == 0) {
+ /* we need to ping to close the window between array
+ * state transitioning to clean and the metadata being
+ * marked clean
+ */
+ rv = 0;
+ } else
+ rv = 1;
+ if (rv && verbose)
+ pr_err("Error waiting for %s to be clean\n",
+ dev);
+
+ /* restore the original safe_mode_delay */
+ sysfs_set_safemode(mdi, mdi->safe_mode_delay);
+ close(state_fd);
+ }
+
+ sysfs_free(mdi);
+ close(fd);
+
+ return rv;
+}
+#endif /* MDASSEMBLE */
projects / thirdparty / mdadm.git / blobdiff