2011-09-13 09:59:48 -04:00
|
|
|
/* Background I/O service for Redis.
|
|
|
|
*
|
|
|
|
* This file implements operations that we need to perform in the background.
|
|
|
|
* Currently there is only a single operation, that is a background close(2)
|
|
|
|
* system call. This is needed as when the process is the last owner of a
|
|
|
|
* reference to a file closing it means unlinking it, and the deletion of the
|
|
|
|
* file is slow, blocking the server.
|
|
|
|
*
|
|
|
|
* In the future we'll either continue implementing new things we need or
|
|
|
|
* we'll switch to libeio. However there are probably long term uses for this
|
|
|
|
* file as we may want to put here Redis specific background tasks (for instance
|
|
|
|
* it is not impossible that we'll need a non blocking FLUSHDB/FLUSHALL
|
|
|
|
* implementation).
|
|
|
|
*
|
|
|
|
* DESIGN
|
|
|
|
* ------
|
|
|
|
*
|
|
|
|
* The design is trivial, we have a structure representing a job to perform
|
2011-09-15 09:46:29 -04:00
|
|
|
* and a different thread and job queue for every job type.
|
|
|
|
* Every thread wait for new jobs in its queue, and process every job
|
|
|
|
* sequentially.
|
|
|
|
*
|
2011-09-15 12:23:11 -04:00
|
|
|
* Jobs of the same type are guaranteed to be processed from the least
|
|
|
|
* recently inserted to the most recently inserted (older jobs processed
|
|
|
|
* first).
|
|
|
|
*
|
2011-09-13 09:59:48 -04:00
|
|
|
* Currently there is no way for the creator of the job to be notified about
|
|
|
|
* the completion of the operation, this will only be added when/if needed.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "redis.h"
|
|
|
|
#include "bio.h"
|
|
|
|
|
2011-09-15 09:46:29 -04:00
|
|
|
static pthread_mutex_t bio_mutex[REDIS_BIO_NUM_OPS];
|
|
|
|
static pthread_cond_t bio_condvar[REDIS_BIO_NUM_OPS];
|
|
|
|
static list *bio_jobs[REDIS_BIO_NUM_OPS];
|
2011-09-14 12:39:29 -04:00
|
|
|
/* The following array is used to hold the number of pending jobs for every
|
|
|
|
* OP type. This allows us to export the bioPendingJobsOfType() API that is
|
|
|
|
* useful when the main thread wants to perform some operation that may involve
|
|
|
|
* objects shared with the background thread. The main thread will just wait
|
|
|
|
* that there are no longer jobs of this type to be executed before performing
|
|
|
|
* the sensible operation. This data is also useful for reporting. */
|
2011-09-15 09:46:29 -04:00
|
|
|
static unsigned long long bio_pending[REDIS_BIO_NUM_OPS];
|
2011-09-13 09:59:48 -04:00
|
|
|
|
|
|
|
/* This structure represents a background Job. It is only used locally to this
|
|
|
|
* file as the API deos not expose the internals at all. */
|
|
|
|
struct bio_job {
|
2011-09-15 09:46:29 -04:00
|
|
|
time_t time; /* Time at which the job was created. */
|
|
|
|
/* Job specific arguments pointers. If we need to pass more than three
|
|
|
|
* arguments we can just pass a pointer to a structure or alike. */
|
|
|
|
void *arg1, *arg2, *arg3;
|
2011-09-13 10:09:06 -04:00
|
|
|
};
|
2011-09-13 09:59:48 -04:00
|
|
|
|
|
|
|
void *bioProcessBackgroundJobs(void *arg);
|
|
|
|
|
2011-09-13 10:09:06 -04:00
|
|
|
/* Make sure we have enough stack to perform all the things we do in the
|
|
|
|
* main thread. */
|
|
|
|
#define REDIS_THREAD_STACK_SIZE (1024*1024*4)
|
|
|
|
|
2011-09-13 09:59:48 -04:00
|
|
|
/* Initialize the background system, spawning the thread. */
|
|
|
|
void bioInit(void) {
|
|
|
|
pthread_attr_t attr;
|
|
|
|
pthread_t thread;
|
|
|
|
size_t stacksize;
|
2011-09-14 12:39:29 -04:00
|
|
|
int j;
|
2011-09-13 09:59:48 -04:00
|
|
|
|
2011-09-14 12:39:29 -04:00
|
|
|
/* Initialization of state vars and objects */
|
2011-09-15 09:46:29 -04:00
|
|
|
for (j = 0; j < REDIS_BIO_NUM_OPS; j++) {
|
|
|
|
pthread_mutex_init(&bio_mutex[j],NULL);
|
|
|
|
pthread_cond_init(&bio_condvar[j],NULL);
|
|
|
|
bio_jobs[j] = listCreate();
|
|
|
|
bio_pending[j] = 0;
|
|
|
|
}
|
2011-09-13 09:59:48 -04:00
|
|
|
|
|
|
|
/* Set the stack size as by default it may be small in some system */
|
|
|
|
pthread_attr_init(&attr);
|
2011-09-13 10:09:06 -04:00
|
|
|
pthread_attr_getstacksize(&attr,&stacksize);
|
2011-09-13 09:59:48 -04:00
|
|
|
if (!stacksize) stacksize = 1; /* The world is full of Solaris Fixes */
|
|
|
|
while (stacksize < REDIS_THREAD_STACK_SIZE) stacksize *= 2;
|
|
|
|
pthread_attr_setstacksize(&attr, stacksize);
|
|
|
|
|
2011-09-15 09:46:29 -04:00
|
|
|
/* Ready to spawn our threads. We use the single argument the thread
|
|
|
|
* function accepts in order to pass the job ID the thread is
|
|
|
|
* responsible of. */
|
|
|
|
for (j = 0; j < REDIS_BIO_NUM_OPS; j++) {
|
|
|
|
void *arg = (void*)(unsigned long) j;
|
|
|
|
if (pthread_create(&thread,&attr,bioProcessBackgroundJobs,arg) != 0) {
|
|
|
|
redisLog(REDIS_WARNING,"Fatal: Can't initialize Background Jobs.");
|
|
|
|
exit(1);
|
|
|
|
}
|
2011-09-13 09:59:48 -04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-09-15 09:46:29 -04:00
|
|
|
void bioCreateBackgroundJob(int type, void *arg1, void *arg2, void *arg3) {
|
2011-09-13 09:59:48 -04:00
|
|
|
struct bio_job *job = zmalloc(sizeof(*job));
|
|
|
|
|
2011-09-15 09:46:29 -04:00
|
|
|
job->time = time(NULL);
|
|
|
|
job->arg1 = arg1;
|
|
|
|
job->arg2 = arg2;
|
|
|
|
job->arg3 = arg3;
|
|
|
|
pthread_mutex_lock(&bio_mutex[type]);
|
|
|
|
listAddNodeTail(bio_jobs[type],job);
|
2011-09-14 12:39:29 -04:00
|
|
|
bio_pending[type]++;
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_cond_signal(&bio_condvar[type]);
|
|
|
|
pthread_mutex_unlock(&bio_mutex[type]);
|
2011-09-13 09:59:48 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
void *bioProcessBackgroundJobs(void *arg) {
|
|
|
|
struct bio_job *job;
|
2011-09-15 09:46:29 -04:00
|
|
|
unsigned long type = (unsigned long) arg;
|
2011-09-13 09:59:48 -04:00
|
|
|
|
|
|
|
pthread_detach(pthread_self());
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_mutex_lock(&bio_mutex[type]);
|
2011-09-13 09:59:48 -04:00
|
|
|
while(1) {
|
|
|
|
listNode *ln;
|
|
|
|
|
|
|
|
/* The loop always starts with the lock hold. */
|
2011-09-15 09:46:29 -04:00
|
|
|
if (listLength(bio_jobs[type]) == 0) {
|
|
|
|
pthread_cond_wait(&bio_condvar[type],&bio_mutex[type]);
|
2011-09-13 09:59:48 -04:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* Pop the job from the queue. */
|
2011-09-15 09:46:29 -04:00
|
|
|
ln = listFirst(bio_jobs[type]);
|
2011-09-13 09:59:48 -04:00
|
|
|
job = ln->value;
|
|
|
|
/* It is now possible to unlock the background system as we know have
|
|
|
|
* a stand alone job structure to process.*/
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_mutex_unlock(&bio_mutex[type]);
|
2011-09-13 09:59:48 -04:00
|
|
|
|
|
|
|
/* Process the job accordingly to its type. */
|
2011-09-14 12:39:29 -04:00
|
|
|
if (type == REDIS_BIO_CLOSE_FILE) {
|
2011-09-15 09:46:29 -04:00
|
|
|
close((long)job->arg1);
|
2011-09-15 12:25:53 -04:00
|
|
|
} else if (type == REDIS_BIO_AOF_FSYNC) {
|
2011-09-16 09:53:01 -04:00
|
|
|
aof_fsync((long)job->arg1);
|
2011-09-13 09:59:48 -04:00
|
|
|
} else {
|
|
|
|
redisPanic("Wrong job type in bioProcessBackgroundJobs().");
|
|
|
|
}
|
|
|
|
zfree(job);
|
|
|
|
|
|
|
|
/* Lock again before reiterating the loop, if there are no longer
|
|
|
|
* jobs to process we'll block again in pthread_cond_wait(). */
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_mutex_lock(&bio_mutex[type]);
|
2011-09-15 09:52:14 -04:00
|
|
|
listDelNode(bio_jobs[type],ln);
|
2011-09-14 12:39:29 -04:00
|
|
|
bio_pending[type]--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Return the number of pending jobs of the specified type. */
|
|
|
|
unsigned long long bioPendingJobsOfType(int type) {
|
|
|
|
unsigned long long val;
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_mutex_lock(&bio_mutex[type]);
|
2011-09-14 12:39:29 -04:00
|
|
|
val = bio_pending[type];
|
2011-09-15 09:46:29 -04:00
|
|
|
pthread_mutex_unlock(&bio_mutex[type]);
|
2011-09-14 12:39:29 -04:00
|
|
|
return val;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Wait until the number of pending jobs of the specified type are
|
|
|
|
* less or equal to the specified number.
|
|
|
|
*
|
|
|
|
* This function may block for long time, it should only be used to perform
|
2011-09-14 13:05:21 -04:00
|
|
|
* the following tasks:
|
|
|
|
*
|
|
|
|
* 1) To avoid that the main thread is pushing jobs of a given time so fast
|
|
|
|
* that the background thread can't process them at the same speed.
|
|
|
|
* So before creating a new job of a given type the main thread should
|
|
|
|
* call something like: bioWaitPendingJobsLE(job_type,10000);
|
|
|
|
* 2) In order to perform special operations that make it necessary to be sure
|
|
|
|
* no one is touching shared resourced in the background.
|
|
|
|
*/
|
2011-09-14 12:39:29 -04:00
|
|
|
void bioWaitPendingJobsLE(int type, unsigned long long num) {
|
|
|
|
unsigned long long iteration = 0;
|
|
|
|
|
|
|
|
/* We poll the jobs queue aggressively to start, and gradually relax
|
|
|
|
* the polling speed if it is going to take too much time. */
|
|
|
|
while(1) {
|
|
|
|
iteration++;
|
|
|
|
if (iteration > 1000 && iteration <= 10000) {
|
|
|
|
usleep(100);
|
|
|
|
} else if (iteration > 10000) {
|
|
|
|
usleep(1000);
|
|
|
|
}
|
|
|
|
if (bioPendingJobsOfType(type) <= num) break;
|
2011-09-13 09:59:48 -04:00
|
|
|
}
|
|
|
|
}
|
2011-09-15 09:46:29 -04:00
|
|
|
|
|
|
|
/* Return the older job of the specified type. */
|
|
|
|
time_t bioOlderJobOfType(int type) {
|
|
|
|
time_t time;
|
|
|
|
listNode *ln;
|
|
|
|
struct bio_job *job;
|
|
|
|
|
|
|
|
pthread_mutex_lock(&bio_mutex[type]);
|
|
|
|
ln = listFirst(bio_jobs[type]);
|
2011-09-15 12:23:58 -04:00
|
|
|
if (ln == NULL) {
|
|
|
|
pthread_mutex_unlock(&bio_mutex[type]);
|
|
|
|
return 0;
|
|
|
|
}
|
2011-09-15 09:46:29 -04:00
|
|
|
job = ln->value;
|
|
|
|
time = job->time;
|
|
|
|
pthread_mutex_unlock(&bio_mutex[type]);
|
|
|
|
return time;
|
|
|
|
}
|
|
|
|
|