Break Monitor into smaller functions.
Monitor() has become way too big. Break it up into multiple smaller functions that are all called from the main loop. Signed-off-by: NeilBrown <neilb@suse.de>
This commit is contained in:
parent
ca750d9830
commit
2e0172b110
457
Monitor.c
457
Monitor.c
|
@ -30,14 +30,44 @@
|
|||
#include <limits.h>
|
||||
#include <syslog.h>
|
||||
|
||||
static void alert(char *event, char *dev, char *disc, char *mailaddr, char *mailfrom,
|
||||
char *cmd, int dosyslog);
|
||||
|
||||
/* 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 */
|
||||
long utime;
|
||||
int err;
|
||||
char *spare_group;
|
||||
int active, working, failed, spare, raid;
|
||||
int expected_spares;
|
||||
int devstate[MaxDisks];
|
||||
unsigned devid[MaxDisks];
|
||||
int percent;
|
||||
int parent_dev; /* For subarray, devnum of parent.
|
||||
* For others, NoMdDev
|
||||
*/
|
||||
struct supertype *metadata;
|
||||
struct state *next;
|
||||
};
|
||||
|
||||
static int make_daemon(char *pidfile);
|
||||
static int check_one_sharer(int scan);
|
||||
static void alert(char *event, char *dev, char *disc, char *mailaddr, char *mailfrom,
|
||||
char *cmd, int dosyslog);
|
||||
static void check_array(struct state *st, struct mdstat_ent *mdstat,
|
||||
int test, char *mailaddr,
|
||||
char *mailfrom, char *alert_cmd, int dosyslog,
|
||||
int increments);
|
||||
static int add_new_arrays(struct mdstat_ent *mdstat, struct state *statelist,
|
||||
int test, char *mailaddr, char *mailfrom,
|
||||
char *alert_cmd, int dosyslog);
|
||||
static void try_spare_migration(struct state *statelist,
|
||||
char *mailaddr, char *mailfrom,
|
||||
char *alert_cmd, int dosyslog);
|
||||
|
||||
int Monitor(struct mddev_dev *devlist,
|
||||
char *mailaddr, char *alert_cmd,
|
||||
int period, int daemonise, int scan, int oneshot,
|
||||
|
@ -86,23 +116,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
* that appears in /proc/mdstat
|
||||
*/
|
||||
|
||||
struct state {
|
||||
char *devname;
|
||||
int devnum; /* to sync with mdstat info */
|
||||
long utime;
|
||||
int err;
|
||||
char *spare_group;
|
||||
int active, working, failed, spare, raid;
|
||||
int expected_spares;
|
||||
int devstate[MaxDisks];
|
||||
unsigned devid[MaxDisks];
|
||||
int percent;
|
||||
int parent_dev; /* For subarray, devnum of parent.
|
||||
* For others, NoMdDev
|
||||
*/
|
||||
struct supertype *metadata;
|
||||
struct state *next;
|
||||
} *statelist = NULL;
|
||||
struct state *statelist = NULL;
|
||||
int finished = 0;
|
||||
struct mdstat_ent *mdstat = NULL;
|
||||
char *mailfrom = NULL;
|
||||
|
@ -126,72 +140,13 @@ int Monitor(struct mddev_dev *devlist,
|
|||
return 1;
|
||||
}
|
||||
|
||||
if (daemonise) {
|
||||
int pid = fork();
|
||||
if (pid > 0) {
|
||||
if (!pidfile)
|
||||
printf("%d\n", pid);
|
||||
else {
|
||||
FILE *pid_file;
|
||||
pid_file=fopen(pidfile, "w");
|
||||
if (!pid_file)
|
||||
perror("cannot create pid file");
|
||||
else {
|
||||
fprintf(pid_file,"%d\n", pid);
|
||||
fclose(pid_file);
|
||||
}
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
if (pid < 0) {
|
||||
perror("daemonise");
|
||||
if (daemonise)
|
||||
if (make_daemon(pidfile))
|
||||
return 1;
|
||||
}
|
||||
close(0);
|
||||
open("/dev/null", O_RDWR);
|
||||
dup2(0,1);
|
||||
dup2(0,2);
|
||||
setsid();
|
||||
}
|
||||
|
||||
if (share) {
|
||||
int pid, rv;
|
||||
FILE *fp;
|
||||
char dir[20];
|
||||
struct stat buf;
|
||||
fp = fopen("/var/run/mdadm/autorebuild.pid", "r");
|
||||
if (fp) {
|
||||
fscanf(fp, "%d", &pid);
|
||||
sprintf(dir, "/proc/%d", pid);
|
||||
rv = stat(dir, &buf);
|
||||
if (rv != -1) {
|
||||
if (scan) {
|
||||
fprintf(stderr, Name ": Only one "
|
||||
"autorebuild process allowed"
|
||||
" in scan mode, aborting\n");
|
||||
fclose(fp);
|
||||
if (share)
|
||||
if (check_one_sharer(scan))
|
||||
return 1;
|
||||
} else {
|
||||
fprintf(stderr, Name ": Warning: One"
|
||||
" autorebuild process already"
|
||||
" running.");
|
||||
}
|
||||
}
|
||||
fclose(fp);
|
||||
}
|
||||
if (scan) {
|
||||
fp = fopen("/var/run/mdadm/autorebuild.pid", "w");
|
||||
if (!fp)
|
||||
fprintf(stderr, Name ": Cannot create"
|
||||
" autorebuild.pid "
|
||||
"file\n");
|
||||
else {
|
||||
pid = getpid();
|
||||
fprintf(fp, "%d\n", pid);
|
||||
fclose(fp);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (devlist == NULL) {
|
||||
struct mddev_ident *mdlist = conf_get_ident(NULL);
|
||||
|
@ -256,7 +211,204 @@ int Monitor(struct mddev_dev *devlist,
|
|||
free_mdstat(mdstat);
|
||||
mdstat = mdstat_read(oneshot?0:1, 0);
|
||||
|
||||
for (st=statelist; st; st=st->next) {
|
||||
for (st=statelist; st; st=st->next)
|
||||
check_array(st, mdstat, test, mailaddr, mailfrom,
|
||||
alert_cmd, dosyslog, increments);
|
||||
|
||||
/* now check if there are any new devices found in mdstat */
|
||||
if (scan)
|
||||
new_found = add_new_arrays(mdstat, statelist, test,
|
||||
mailaddr, mailfrom, alert_cmd,
|
||||
dosyslog);
|
||||
|
||||
/* If an array has active < raid && spare == 0 && spare_group != NULL
|
||||
* Look for another array with spare > 0 and active == raid and same spare_group
|
||||
* if found, choose a device and hotremove/hotadd
|
||||
*/
|
||||
if (share)
|
||||
try_spare_migration(statelist, mailaddr, mailfrom,
|
||||
alert_cmd, dosyslog);
|
||||
if (!new_found) {
|
||||
if (oneshot)
|
||||
break;
|
||||
else
|
||||
mdstat_wait(period);
|
||||
}
|
||||
test = 0;
|
||||
}
|
||||
if (pidfile)
|
||||
unlink(pidfile);
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int make_daemon(char *pidfile)
|
||||
{
|
||||
int pid = fork();
|
||||
if (pid > 0) {
|
||||
if (!pidfile)
|
||||
printf("%d\n", pid);
|
||||
else {
|
||||
FILE *pid_file;
|
||||
pid_file=fopen(pidfile, "w");
|
||||
if (!pid_file)
|
||||
perror("cannot create pid file");
|
||||
else {
|
||||
fprintf(pid_file,"%d\n", pid);
|
||||
fclose(pid_file);
|
||||
}
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
if (pid < 0) {
|
||||
perror("daemonise");
|
||||
return 1;
|
||||
}
|
||||
close(0);
|
||||
open("/dev/null", O_RDWR);
|
||||
dup2(0,1);
|
||||
dup2(0,2);
|
||||
setsid();
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int check_one_sharer(int scan)
|
||||
{
|
||||
int pid, rv;
|
||||
FILE *fp;
|
||||
char dir[20];
|
||||
struct stat buf;
|
||||
fp = fopen("/var/run/mdadm/autorebuild.pid", "r");
|
||||
if (fp) {
|
||||
fscanf(fp, "%d", &pid);
|
||||
sprintf(dir, "/proc/%d", pid);
|
||||
rv = stat(dir, &buf);
|
||||
if (rv != -1) {
|
||||
if (scan) {
|
||||
fprintf(stderr, Name ": Only one "
|
||||
"autorebuild process allowed"
|
||||
" in scan mode, aborting\n");
|
||||
fclose(fp);
|
||||
return 1;
|
||||
} else {
|
||||
fprintf(stderr, Name ": Warning: One"
|
||||
" autorebuild process already"
|
||||
" running.");
|
||||
}
|
||||
}
|
||||
fclose(fp);
|
||||
}
|
||||
if (scan) {
|
||||
fp = fopen("/var/run/mdadm/autorebuild.pid", "w");
|
||||
if (!fp)
|
||||
fprintf(stderr, Name ": Cannot create"
|
||||
" autorebuild.pid "
|
||||
"file\n");
|
||||
else {
|
||||
pid = getpid();
|
||||
fprintf(fp, "%d\n", pid);
|
||||
fclose(fp);
|
||||
}
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
static void alert(char *event, char *dev, char *disc, char *mailaddr, char *mailfrom, char *cmd,
|
||||
int dosyslog)
|
||||
{
|
||||
int priority;
|
||||
|
||||
if (!cmd && !mailaddr) {
|
||||
time_t now = time(0);
|
||||
|
||||
printf("%1.15s: %s on %s %s\n", ctime(&now)+4, event, dev, disc?disc:"unknown device");
|
||||
}
|
||||
if (cmd) {
|
||||
int pid = fork();
|
||||
switch(pid) {
|
||||
default:
|
||||
waitpid(pid, NULL, 0);
|
||||
break;
|
||||
case -1:
|
||||
break;
|
||||
case 0:
|
||||
execl(cmd, cmd, event, dev, disc, NULL);
|
||||
exit(2);
|
||||
}
|
||||
}
|
||||
if (mailaddr &&
|
||||
(strncmp(event, "Fail", 4)==0 ||
|
||||
strncmp(event, "Test", 4)==0 ||
|
||||
strncmp(event, "Spares", 6)==0 ||
|
||||
strncmp(event, "Degrade", 7)==0)) {
|
||||
FILE *mp = popen(Sendmail, "w");
|
||||
if (mp) {
|
||||
FILE *mdstat;
|
||||
char hname[256];
|
||||
gethostname(hname, sizeof(hname));
|
||||
signal(SIGPIPE, SIG_IGN);
|
||||
if (mailfrom)
|
||||
fprintf(mp, "From: %s\n", mailfrom);
|
||||
else
|
||||
fprintf(mp, "From: " Name " monitoring <root>\n");
|
||||
fprintf(mp, "To: %s\n", mailaddr);
|
||||
fprintf(mp, "Subject: %s event on %s:%s\n\n", event, dev, hname);
|
||||
|
||||
fprintf(mp, "This is an automatically generated mail message from " Name "\n");
|
||||
fprintf(mp, "running on %s\n\n", hname);
|
||||
|
||||
fprintf(mp, "A %s event had been detected on md device %s.\n\n", event, dev);
|
||||
|
||||
if (disc && disc[0] != ' ')
|
||||
fprintf(mp, "It could be related to component device %s.\n\n", disc);
|
||||
if (disc && disc[0] == ' ')
|
||||
fprintf(mp, "Extra information:%s.\n\n", disc);
|
||||
|
||||
fprintf(mp, "Faithfully yours, etc.\n");
|
||||
|
||||
mdstat = fopen("/proc/mdstat", "r");
|
||||
if (mdstat) {
|
||||
char buf[8192];
|
||||
int n;
|
||||
fprintf(mp, "\nP.S. The /proc/mdstat file currently contains the following:\n\n");
|
||||
while ( (n=fread(buf, 1, sizeof(buf), mdstat)) > 0)
|
||||
n=fwrite(buf, 1, n, mp); /* yes, i don't care about the result */
|
||||
fclose(mdstat);
|
||||
}
|
||||
pclose(mp);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
/* log the event to syslog maybe */
|
||||
if (dosyslog) {
|
||||
/* Log at a different severity depending on the event.
|
||||
*
|
||||
* These are the critical events: */
|
||||
if (strncmp(event, "Fail", 4)==0 ||
|
||||
strncmp(event, "Degrade", 7)==0 ||
|
||||
strncmp(event, "DeviceDisappeared", 17)==0)
|
||||
priority = LOG_CRIT;
|
||||
/* Good to know about, but are not failures: */
|
||||
else if (strncmp(event, "Rebuild", 7)==0 ||
|
||||
strncmp(event, "MoveSpare", 9)==0 ||
|
||||
strncmp(event, "Spares", 6) != 0)
|
||||
priority = LOG_WARNING;
|
||||
/* Everything else: */
|
||||
else
|
||||
priority = LOG_INFO;
|
||||
|
||||
if (disc)
|
||||
syslog(priority, "%s event detected on md device %s, component device %s", event, dev, disc);
|
||||
else
|
||||
syslog(priority, "%s event detected on md device %s", event, dev);
|
||||
}
|
||||
}
|
||||
|
||||
static void check_array(struct state *st, struct mdstat_ent *mdstat,
|
||||
int test, char *mailaddr,
|
||||
char *mailfrom, char *alert_cmd, int dosyslog,
|
||||
int increments)
|
||||
{
|
||||
struct { int state, major, minor; } info[MaxDisks];
|
||||
mdu_array_info_t array;
|
||||
struct mdstat_ent *mse = NULL, *mse2;
|
||||
|
@ -274,7 +426,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
/* fprintf(stderr, Name ": cannot open %s: %s\n",
|
||||
dev, strerror(errno));
|
||||
*/ st->err=1;
|
||||
continue;
|
||||
return;
|
||||
}
|
||||
fcntl(fd, F_SETFD, FD_CLOEXEC);
|
||||
if (ioctl(fd, GET_ARRAY_INFO, &array)<0) {
|
||||
|
@ -285,7 +437,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
dev, strerror(errno));
|
||||
*/ st->err=1;
|
||||
close(fd);
|
||||
continue;
|
||||
return;
|
||||
}
|
||||
/* It's much easier to list what array levels can't
|
||||
* have a device disappear than all of them that can
|
||||
|
@ -296,7 +448,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
mailaddr, mailfrom, alert_cmd, dosyslog);
|
||||
st->err = 1;
|
||||
close(fd);
|
||||
continue;
|
||||
return;
|
||||
}
|
||||
if (st->devnum == INT_MAX) {
|
||||
struct stat stb;
|
||||
|
@ -320,7 +472,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
* or re-created after reading mdstat*/
|
||||
st->err = 1;
|
||||
close(fd);
|
||||
continue;
|
||||
return;
|
||||
}
|
||||
/* this array is in /proc/mdstat */
|
||||
if (array.utime == 0)
|
||||
|
@ -337,7 +489,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
))) {
|
||||
close(fd);
|
||||
st->err = 0;
|
||||
continue;
|
||||
return;
|
||||
}
|
||||
if (st->utime == 0 && /* new array */
|
||||
mse->pattern && strchr(mse->pattern, '_') /* degraded */
|
||||
|
@ -466,9 +618,14 @@ int Monitor(struct mddev_dev *devlist,
|
|||
st->raid = array.raid_disks;
|
||||
st->err = 0;
|
||||
}
|
||||
/* now check if there are any new devices found in mdstat */
|
||||
if (scan) {
|
||||
|
||||
static int add_new_arrays(struct mdstat_ent *mdstat, struct state *statelist,
|
||||
int test, char *mailaddr, char *mailfrom,
|
||||
char *alert_cmd, int dosyslog)
|
||||
{
|
||||
struct mdstat_ent *mse;
|
||||
int new_found = 0;
|
||||
|
||||
for (mse=mdstat; mse; mse=mse->next)
|
||||
if (mse->devnum != INT_MAX &&
|
||||
(!mse->level || /* retrieve containers */
|
||||
|
@ -515,12 +672,15 @@ int Monitor(struct mddev_dev *devlist,
|
|||
alert("NewArray", st->devname, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
|
||||
new_found = 1;
|
||||
}
|
||||
return new_found;
|
||||
}
|
||||
/* If an array has active < raid && spare == 0 && spare_group != NULL
|
||||
* Look for another array with spare > 0 and active == raid and same spare_group
|
||||
* if found, choose a device and hotremove/hotadd
|
||||
*/
|
||||
if (share) for (st = statelist; st; st=st->next)
|
||||
|
||||
static void try_spare_migration(struct state *statelist,
|
||||
char *mailaddr, char *mailfrom,
|
||||
char *alert_cmd, int dosyslog)
|
||||
{
|
||||
struct state *st;
|
||||
for (st = statelist; st; st=st->next)
|
||||
if (st->active < st->raid &&
|
||||
st->spare == 0 &&
|
||||
st->spare_group != NULL) {
|
||||
|
@ -574,112 +734,7 @@ int Monitor(struct mddev_dev *devlist,
|
|||
close(fd2);
|
||||
}
|
||||
}
|
||||
if (!new_found) {
|
||||
if (oneshot)
|
||||
break;
|
||||
else
|
||||
mdstat_wait(period);
|
||||
}
|
||||
test = 0;
|
||||
}
|
||||
if (pidfile)
|
||||
unlink(pidfile);
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
||||
static void alert(char *event, char *dev, char *disc, char *mailaddr, char *mailfrom, char *cmd,
|
||||
int dosyslog)
|
||||
{
|
||||
int priority;
|
||||
|
||||
if (!cmd && !mailaddr) {
|
||||
time_t now = time(0);
|
||||
|
||||
printf("%1.15s: %s on %s %s\n", ctime(&now)+4, event, dev, disc?disc:"unknown device");
|
||||
}
|
||||
if (cmd) {
|
||||
int pid = fork();
|
||||
switch(pid) {
|
||||
default:
|
||||
waitpid(pid, NULL, 0);
|
||||
break;
|
||||
case -1:
|
||||
break;
|
||||
case 0:
|
||||
execl(cmd, cmd, event, dev, disc, NULL);
|
||||
exit(2);
|
||||
}
|
||||
}
|
||||
if (mailaddr &&
|
||||
(strncmp(event, "Fail", 4)==0 ||
|
||||
strncmp(event, "Test", 4)==0 ||
|
||||
strncmp(event, "Spares", 6)==0 ||
|
||||
strncmp(event, "Degrade", 7)==0)) {
|
||||
FILE *mp = popen(Sendmail, "w");
|
||||
if (mp) {
|
||||
FILE *mdstat;
|
||||
char hname[256];
|
||||
gethostname(hname, sizeof(hname));
|
||||
signal(SIGPIPE, SIG_IGN);
|
||||
if (mailfrom)
|
||||
fprintf(mp, "From: %s\n", mailfrom);
|
||||
else
|
||||
fprintf(mp, "From: " Name " monitoring <root>\n");
|
||||
fprintf(mp, "To: %s\n", mailaddr);
|
||||
fprintf(mp, "Subject: %s event on %s:%s\n\n", event, dev, hname);
|
||||
|
||||
fprintf(mp, "This is an automatically generated mail message from " Name "\n");
|
||||
fprintf(mp, "running on %s\n\n", hname);
|
||||
|
||||
fprintf(mp, "A %s event had been detected on md device %s.\n\n", event, dev);
|
||||
|
||||
if (disc && disc[0] != ' ')
|
||||
fprintf(mp, "It could be related to component device %s.\n\n", disc);
|
||||
if (disc && disc[0] == ' ')
|
||||
fprintf(mp, "Extra information:%s.\n\n", disc);
|
||||
|
||||
fprintf(mp, "Faithfully yours, etc.\n");
|
||||
|
||||
mdstat = fopen("/proc/mdstat", "r");
|
||||
if (mdstat) {
|
||||
char buf[8192];
|
||||
int n;
|
||||
fprintf(mp, "\nP.S. The /proc/mdstat file currently contains the following:\n\n");
|
||||
while ( (n=fread(buf, 1, sizeof(buf), mdstat)) > 0)
|
||||
n=fwrite(buf, 1, n, mp); /* yes, i don't care about the result */
|
||||
fclose(mdstat);
|
||||
}
|
||||
pclose(mp);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
/* log the event to syslog maybe */
|
||||
if (dosyslog) {
|
||||
/* Log at a different severity depending on the event.
|
||||
*
|
||||
* These are the critical events: */
|
||||
if (strncmp(event, "Fail", 4)==0 ||
|
||||
strncmp(event, "Degrade", 7)==0 ||
|
||||
strncmp(event, "DeviceDisappeared", 17)==0)
|
||||
priority = LOG_CRIT;
|
||||
/* Good to know about, but are not failures: */
|
||||
else if (strncmp(event, "Rebuild", 7)==0 ||
|
||||
strncmp(event, "MoveSpare", 9)==0 ||
|
||||
strncmp(event, "Spares", 6) != 0)
|
||||
priority = LOG_WARNING;
|
||||
/* Everything else: */
|
||||
else
|
||||
priority = LOG_INFO;
|
||||
|
||||
if (disc)
|
||||
syslog(priority, "%s event detected on md device %s, component device %s", event, dev, disc);
|
||||
else
|
||||
syslog(priority, "%s event detected on md device %s", event, dev);
|
||||
}
|
||||
}
|
||||
|
||||
/* Not really Monitor but ... */
|
||||
int Wait(char *dev)
|
||||
{
|
||||
|
|
Loading…
Reference in New Issue