lei.chen/myxce
1
0
Fork 1
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

save works

Browse source
This commit is contained in:
Rongsong Shen 2026-03-02 12:14:30 +08:00
parent b748e30d20
commit f4b1ba4cd7
21 changed files with 373 additions and 324 deletions
Download patch file Download diff file Expand all files Collapse all files

16
src/controller/apiserver.c
View file

@ -256,7 +256,7 @@ static int apiserver_job_update_fun(int type, int situation, void* data,
int rc;
xce_notify_data_t xdata
= { .event = NotifyData_data_updateJob,
.data = { .update_job = job->job } };
.data = { .update_job = &(job->job) } };
rc = notify(c, c->apiservers[i].ep, &xdata);
if (rc < 0) {
@ -351,11 +351,11 @@ static int apiserver_worker_change(int type, int situation, void* data,
{
int i;
TRACE("%s worker %s idx %d situation %d", __func__, w->host->name, w->idx,
TRACE("%s worker %s idx %d situation %d", __func__, w->host.name, w->idx,
situation);
collect_resval_with_idx(&(w->host->resvals), &(w->host->nresvals_to_sched),
&(w->host->resvals_to_sched));
collect_resval_with_idx(&(w->host.resvals), &(w->host.nresvals_to_sched),
&(w->host.resvals_to_sched));
switch (situation) {
case HOOK_AFTER_EVENT:
@ -383,8 +383,8 @@ static int apiserver_worker_change(int type, int situation, void* data,
break;
}
free_resval_with_idx_list(&(w->host->resvals_to_sched),
w->host->nresvals_to_sched);
free_resval_with_idx_list(&(w->host.resvals_to_sched),
w->host.nresvals_to_sched);
return 0;
}
@ -402,7 +402,7 @@ static int apiserver_worker_join(int type, int situation, void* data,
return 0;
}
rc = key_find(c->workermgr, (void *)name, (void **)&w);
rc = key_find(c->workermgr, (void*)name, (void**)&w);
if (rc < 0) {
return rc;
}
@ -514,7 +514,7 @@ static int apiserver_queue_hook_fun(int type, int situation, int detail,
int rc;
xce_notify_data_t xdata
= { .event = NotifyData_data_schedUpdateQueue,
.data = { .sched_update_queue = q->queue } };
.data = { .sched_update_queue = &(q->queue) } };
rc = notify(c, c->apiservers[i].ep, &xdata);
if (rc < 0) {

26
src/controller/cluster.c
View file

@ -515,26 +515,18 @@ int cluster_init(cluster_t* c)
kvmgr_params_t resmgr_params
= { .key_type = KVMGR_KEY_STRING,
.dynamic = 1,
.nslots = c->myconf->cluster_config->resmgr->nhash_slots,
.delta = c->myconf->cluster_config->resmgr->delta,
.datakiller = (destroy_fn_t)res_obj_killer };
kvmgr_params_t jobmgr_params
= { .key_type = KVMGR_KEY_INT,
.dynamic = 1,
.nslots = c->myconf->cluster_config->jobmgr->nhash_slots,
.delta = c->myconf->cluster_config->jobmgr->delta,
.datakiller = (destroy_fn_t)job_obj_killer };
kvmgr_params_t workermgr_params
= { .key_type = KVMGR_KEY_STRING,
.dynamic = 1,
.nslots = c->myconf->cluster_config->workermgr->nhash_slots,
.delta = c->myconf->cluster_config->workermgr->delta,
.datakiller = (destroy_fn_t)worker_obj_killer };
kvmgr_params_t qmgr_params
= { .key_type = KVMGR_KEY_STRING,
.dynamic = 0,
.nslots = c->myconf->cluster_config->qmgr->nhash_slots,
.delta = c->myconf->cluster_config->qmgr->delta,
.datakiller = (destroy_fn_t)queue_obj_killer };
if (c == NULL) {
@ -565,6 +557,15 @@ int cluster_init(cluster_t* c)
return -errno;
}
resmgr_params.nslots = c->myconf->cluster_config->resmgr->nhash_slots;
resmgr_params.delta = c->myconf->cluster_config->resmgr->delta;
jobmgr_params.nslots = c->myconf->cluster_config->jobmgr->nhash_slots;
jobmgr_params.delta = c->myconf->cluster_config->jobmgr->delta;
workermgr_params.nslots = c->myconf->cluster_config->workermgr->nhash_slots;
workermgr_params.delta = c->myconf->cluster_config->workermgr->delta;
qmgr_params.nslots = c->myconf->cluster_config->qmgr->nhash_slots;
qmgr_params.delta = c->myconf->cluster_config->qmgr->delta;
c->configmgr = makeInstance(&xkvmgr_type, &config_mgr_params);
objmgr_init(&(c->objmgr));
@ -644,13 +645,13 @@ static int check_worker_timeout_(cluster_t* c, worker_t* w)
if (now - ep->ts_last_in > timeout) {
TRACE("%s worker %s becomes UNAVAIL as heartbeat timeout", __func__,
w->host->name);
w->host.name);
mark_worker_unavail(c, w);
return 0;
} else {
if (IS_WORKER_UNAVAIL(w->host->status)) {
if (IS_WORKER_UNAVAIL(w->host.status)) {
mark_worker_back(c, w);
}
}
@ -659,13 +660,12 @@ static int check_worker_timeout_(cluster_t* c, worker_t* w)
if (DATAQ_EMPTY(&(ep->outq))) {
int rc = 0;
TRACE("%s send heartbeat to %s", __func__, w->host->name);
TRACE("%s send heartbeat to %s", __func__, w->host.name);
rc = make_heartbeat(c, ep);
if (rc < 0) {
ERROR("%s make_heartbeat to %s failed", __func__,
w->host->name);
ERROR("%s make_heartbeat to %s failed", __func__, w->host.name);
return rc;
}

143
src/controller/job.c
View file

@ -23,13 +23,7 @@ static job_t* create_job(cluster_t* c, xce_job_t* j)
{
job_t* job = NULL;
job = calloc(1, sizeof(*job));
if (job == NULL) {
return NULL;
}
/* FIXME: */
job->job = makeJob(j);
job = makeJob(c->jobmgr, j);
return job;
}
@ -198,9 +192,9 @@ static int dispatch_a_job_(cluster_t* c, endpoint_t* ep, xce_job_alloc_t* alloc)
return -errno;
}
if ((w->ep == NULL) || (w->host->status != 0)) {
if ((w->ep == NULL) || (w->host.status != 0)) {
ERROR("worker %s is not in correct state, status %x ep %p",
w->host->name, w->host->status, w->ep);
w->host.name, w->host.status, w->ep);
free_launchinfo(&launchinfo);
obj_deref(w);
@ -215,41 +209,40 @@ static int dispatch_a_job_(cluster_t* c, endpoint_t* ep, xce_job_alloc_t* alloc)
char* q_pre_exec;
char* q_post_exec;
rc = get_queue_exec(c, j->job->submit->queue, &q_pre_exec,
&q_post_exec);
rc = get_queue_exec(c, j->job.submit->queue, &q_pre_exec, &q_post_exec);
if (rc < 0) {
obj_deref(j);
obj_deref(w);
return rc;
}
if (j->job->submit->start_info->q_pre_exec != NULL) {
obj_deref(j->job->submit->start_info->q_pre_exec);
if (j->job.submit->start_info->q_pre_exec != NULL) {
obj_deref(j->job.submit->start_info->q_pre_exec);
}
if (j->job->submit->start_info->q_post_exec != NULL) {
obj_deref(j->job->submit->start_info->q_post_exec);
if (j->job.submit->start_info->q_post_exec != NULL) {
obj_deref(j->job.submit->start_info->q_post_exec);
}
j->job->submit->start_info->q_pre_exec = q_pre_exec;
j->job->submit->start_info->q_post_exec = q_post_exec;
j->job.submit->start_info->q_pre_exec = q_pre_exec;
j->job.submit->start_info->q_post_exec = q_post_exec;
}
einfo.jobid = alloc->jobid;
einfo.start_info = j->job->submit->start_info;
einfo.start_info = j->job.submit->start_info;
einfo.launchinfo = &launchinfo;
rc = xce_startjob_req(w->ep, &einfo);
free_launchinfo(&launchinfo);
if (j->job->allocs != NULL) {
free_JobAlloc_ptr(&(j->job->allocs));
if (j->job.allocs != NULL) {
free_JobAlloc_ptr(&(j->job.allocs));
}
j->job->allocs = (xce_job_alloc_t*)calloc(1, sizeof(xce_job_alloc_t));
j->job->allocs->jobid = alloc->jobid;
j->job->allocs->nhosts = nc;
MOVE_PTR(j->job->allocs->allocs, allocs);
j->job.allocs = (xce_job_alloc_t*)calloc(1, sizeof(xce_job_alloc_t));
j->job.allocs->jobid = alloc->jobid;
j->job.allocs->nhosts = nc;
MOVE_PTR(j->job.allocs->allocs, allocs);
alloc->allocs = NULL;
DEREF_OBJ(w);
@ -282,8 +275,8 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
return -ENOENT;
}
if (j->job->allocs != NULL) {
xce_host_alloc_t** allocs = j->job->allocs->allocs;
if (j->job.allocs != NULL) {
xce_host_alloc_t** allocs = j->job.allocs->allocs;
worker_t* w = NULL;
/* job control command will be sent to first node
@ -295,15 +288,15 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
}
if (w != NULL) {
/* FIXME: better method to check host status */
if ((w->ep != NULL) && (w->host->status == 0)) {
if ((w->ep != NULL) && (w->host.status == 0)) {
rc = xce_jobctrl_req(w->ep, ctrinfo);
if (rc < 0) {
ERROR("job ctrl %ld on host %s returns %d", jobid,
w->host->name, rc);
w->host.name, rc);
}
} else {
ERROR("worker %s is not in correct state, status %x ep %p",
w->host->name, w->host->status, w->ep);
w->host.name, w->host.status, w->ep);
rc = -errno;
}
obj_deref(w);
@ -322,9 +315,9 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
case JobControl_u_siginfo:
switch (ctrinfo->u.siginfo->signo) {
case SIGCONT:
if (j->job->state
if (j->job.state
& (JOB_STAT_PSUSP | JOB_STAT_USUSP | JOB_STAT_SSUSP)) {
int state = j->job->state;
int state = j->job.state;
/* resume suspend job? */
state &= ~(JOB_STAT_PSUSP | JOB_STAT_USUSP);
@ -333,9 +326,9 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
break;
case SIGSTOP:
case SIGTSTP:
if (!(j->job->state
if (!(j->job.state
& (JOB_STAT_PSUSP | JOB_STAT_USUSP | JOB_STAT_SSUSP))) {
int state = j->job->state;
int state = j->job.state;
state |= JOB_STAT_PSUSP;
mark_job_update(c, j, state);
@ -344,7 +337,7 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
case SIGKILL:
case SIGTERM:
default:
if (j->job->state
if (j->job.state
& (JOB_STAT_DONE | JOB_STAT_EXIT | JOB_STAT_ERR)) {
break;
}
@ -357,8 +350,7 @@ int job_ctrl(cluster_t* c, xce_jobctrl_t* ctrinfo)
}
break;
case JobControl_u_ckptoption:
if (j->job->state
& (JOB_STAT_DONE | JOB_STAT_EXIT | JOB_STAT_ERR)) {
if (j->job.state & (JOB_STAT_DONE | JOB_STAT_EXIT | JOB_STAT_ERR)) {
ERROR("job <%ld> has been finished, no checkpoint can be done.",
jobid);
} else {
@ -509,20 +501,20 @@ static int mark_job_start(cluster_t* c, job_t* job, job_pidinfo_t* info)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_JOB_START, c, info, job);
/* update job information */
job->job->state = JOB_STAT_RUN;
job->job->start_time = time(NULL);
if (job->job->pidinfo == NULL) {
job->job->pidinfo = (job_pidinfo_t*)calloc(1, sizeof(job_pidinfo_t));
job->job.state = JOB_STAT_RUN;
job->job.start_time = time(NULL);
if (job->job.pidinfo == NULL) {
job->job.pidinfo = (job_pidinfo_t*)calloc(1, sizeof(job_pidinfo_t));
}
job->job->pidinfo->npids = copy_int_list((int**)&(job->job->pidinfo->pids),
(int*)info->pids, info->npids);
job->job->pidinfo->npgids = copy_int_list(
(int**)&(job->job->pidinfo->pgids), (int*)info->pgids, info->npgids);
job->job.pidinfo->npids = copy_int_list((int**)&(job->job.pidinfo->pids),
(int*)info->pids, info->npids);
job->job.pidinfo->npgids = copy_int_list((int**)&(job->job.pidinfo->pgids),
(int*)info->pgids, info->npgids);
if (info->exdata != NULL) {
job->job->pidinfo->exdata = makeString(info->exdata);
job->job.pidinfo->exdata = makeString(info->exdata);
} else {
job->job->pidinfo->exdata = NULL;
job->job.pidinfo->exdata = NULL;
}
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_AFTER_EVENT,
@ -556,24 +548,23 @@ static int mark_job_finish(cluster_t* c, job_t* job, xce_jobfinish_t* finish)
HOOK_DETAIL_JOB_EXIT, c, finish, job);
/* update job information */
if (finish->exit_code == 0) {
job->job->state = JOB_STAT_DONE;
job->job.state = JOB_STAT_DONE;
} else {
job->job->state = JOB_STAT_EXIT;
job->job.state = JOB_STAT_EXIT;
}
job->job->finish_time = time(NULL);
job->job.finish_time = time(NULL);
/* fill job run time/rusage information */
if (job->job->exit_info == NULL) {
job->job->exit_info
= (job_exitinfo_t*)calloc(1, sizeof(job_exitinfo_t));
if (job->job.exit_info == NULL) {
job->job.exit_info = (job_exitinfo_t*)calloc(1, sizeof(job_exitinfo_t));
}
if (job->job->exit_info != NULL) {
job->job->exit_info->exitcode = finish->exit_code;
if (job->job.exit_info != NULL) {
job->job.exit_info->exitcode = finish->exit_code;
}
job->job->run_time = job->job->finish_time - job->job->start_time;
job->job.run_time = job->job.finish_time - job->job.start_time;
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_AFTER_EVENT,
HOOK_DETAIL_JOB_EXIT, c, finish, job);
@ -609,20 +600,20 @@ static int mark_job_fail(cluster_t* c, job_t* job, job_failinfo_t* failinfo)
switch (failinfo->stage) {
case JOB_EXEC_STAGE_Q_POST_EXEC:
case JOB_EXEC_STAGE_POST_EXEC:
if ((job->job->state & (JOB_STAT_EXIT | JOB_STAT_DONE)) == 0) {
if ((job->job.state & (JOB_STAT_EXIT | JOB_STAT_DONE)) == 0) {
WARN("job %ld got post exec fail notification but job's exit "
"status has not been set",
job->job->jobid);
job->job->state |= JOB_STAT_EXIT;
job->job.jobid);
job->job.state |= JOB_STAT_EXIT;
}
job->job->state |= JOB_STAT_PERR;
job->job.state |= JOB_STAT_PERR;
break;
default:
job->job->state = JOB_STAT_EXIT;
job->job.state = JOB_STAT_EXIT;
break;
}
job->job->finish_time = time(NULL);
job->job.finish_time = time(NULL);
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_AFTER_EVENT,
HOOK_DETAIL_JOB_FAIL, c, failinfo, job);
@ -635,7 +626,7 @@ static int mark_job_update(cluster_t* c, job_t* job, int state)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_JOB_UPDATE, c, job);
/* update job state */
job->job->state = state;
job->job.state = state;
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_AFTER_EVENT,
HOOK_DETAIL_JOB_UPDATE, c, job);
@ -674,7 +665,7 @@ int on_job_stop(cluster_t* c, uint64_t jobid)
return -errno;
}
state = job->job->state;
state = job->job.state;
state |= JOB_STAT_USUSP;
mark_job_update(c, job, state);
@ -695,7 +686,7 @@ int on_job_resume(cluster_t* c, uint64_t jobid)
return -errno;
}
state = job->job->state;
state = job->job.state;
state &= ~JOB_STAT_USUSP;
mark_job_update(c, job, state);
@ -715,14 +706,14 @@ int on_job_postdone(cluster_t* c, endpoint_t* ep, uint64_t jobid)
return -errno;
}
state = job->job->state;
state = job->job.state;
if ((state & (JOB_STAT_EXIT | JOB_STAT_DONE)) == 0) {
WARN("job %ld got post exec fail notification but job's exit status "
"has not been set",
job->job->jobid);
job->job.jobid);
state |= JOB_STAT_EXIT;
/* ensure that job finish time has been set */
job->job->finish_time = time(NULL);
job->job.finish_time = time(NULL);
}
state |= JOB_STAT_PDONE;
@ -756,11 +747,11 @@ int on_job_ckpt(cluster_t* c, endpoint_t* ep, xce_ckptinfo_t* ckptinfo)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_JOB_CKPT, c, j);
if (j->job->ckptinfo != NULL) {
obj_deref(j->job->ckptinfo);
if (j->job.ckptinfo != NULL) {
obj_deref(j->job.ckptinfo);
}
j->job->ckptinfo = makeString(info);
j->job.ckptinfo = makeString(info);
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_AFTER_EVENT,
HOOK_DETAIL_JOB_CKPT, c, j);
@ -809,7 +800,7 @@ int update_job_rusages(cluster_t* c, endpoint_t* ep, xce_request_t* request)
return errno;
}
host = makeString(w->host->name);
host = makeString(w->host.name);
obj_deref(w);
if (host == NULL) {
@ -821,12 +812,12 @@ int update_job_rusages(cluster_t* c, endpoint_t* ep, xce_request_t* request)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_JOB_UPDATE, c, j);
update_host_runusage(&(j->job->run_usage), host, rusages->count,
update_host_runusage(&(j->job.run_usage), host, rusages->count,
rusages->usages);
update_pids(&(j->job->pidinfo), rusages->npids, rusages->pids);
update_pids(&(j->job.pidinfo), rusages->npids, rusages->pids);
update_pgids(&(j->job->pidinfo), rusages->npgids, rusages->pgids);
update_pgids(&(j->job.pidinfo), rusages->npgids, rusages->pgids);
/* FIXME
* update other fields of job rusage later
@ -856,7 +847,7 @@ static int set_preason(cluster_t* c, xce_pendreason_t* reasons)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_SETPREASON, c, j);
job = j->job;
job = &(j->job);
if (job->pend_reasons != NULL) {
erase_job_preason(job->pend_reasons);
FREEUP(job->pend_reasons);
@ -923,7 +914,7 @@ int do_clr_preason(cluster_t* c, endpoint_t* ep, xce_request_t* request)
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_JOB, HOOK_BEFORE_EVENT,
HOOK_DETAIL_CLRPREASON, c, j);
job = j->job;
job = &(j->job);
if (job->pend_reasons != NULL) {
erase_job_preason(job->pend_reasons);
FREEUP(job->pend_reasons);

2
src/controller/opget.c
View file

@ -157,7 +157,7 @@ static int do_reply_joblist(cluster_t* c, endpoint_t* ep,
if (rc > 0) {
jobdata.jobs = (xce_job_t**)calloc(1, sizeof(xce_job_t*));
if (jobdata.jobs != NULL) {
jobdata.jobs[0] = j->job;
jobdata.jobs[0] = &(j->job);
jobdata.n_total = 1;
}
}

31
src/controller/queue.c
View file

@ -19,12 +19,7 @@ queue_t* create_queue(cluster_t* c, xce_queue_t* p)
{
queue_t* q;
q = calloc(1, sizeof(*q));
if (q == NULL) {
return NULL;
}
q->queue = makeQueue(p);
q = makeQueue(c->qmgr, p);
return q;
}
@ -72,28 +67,28 @@ int queue_ctrl(cluster_t* c, xce_queuectrl_t* ctrl)
switch (ctrl->action) {
case QUEUE_OPEN:
if (!(q->queue->status & QUEUE_STAT_CLOSED)) {
if (!(q->queue.status & QUEUE_STAT_CLOSED)) {
goto clean;
}
q->queue->status &= ~QUEUE_STAT_CLOSED;
q->queue.status &= ~QUEUE_STAT_CLOSED;
break;
case QUEUE_CLOSE:
if (q->queue->status & QUEUE_STAT_CLOSED) {
if (q->queue.status & QUEUE_STAT_CLOSED) {
goto clean;
}
q->queue->status |= QUEUE_STAT_CLOSED;
q->queue.status |= QUEUE_STAT_CLOSED;
break;
case QUEUE_INACTIVE:
if (q->queue->status & QUEUE_STAT_INACT_ADMIN) {
if (q->queue.status & QUEUE_STAT_INACT_ADMIN) {
goto clean;
}
q->queue->status |= QUEUE_STAT_INACT_ADMIN;
q->queue.status |= QUEUE_STAT_INACT_ADMIN;
break;
case QUEUE_ACTIVE:
if (!(q->queue->status & QUEUE_STAT_INACT_ADMIN)) {
if (!(q->queue.status & QUEUE_STAT_INACT_ADMIN)) {
goto clean;
}
q->queue->status &= ~QUEUE_STAT_INACT_ADMIN;
q->queue.status &= ~QUEUE_STAT_INACT_ADMIN;
break;
default:
rc = -1;
@ -147,14 +142,14 @@ int get_queue_exec(cluster_t* c, char* queue, char** pre_exec, char** post_exec)
return -ENOENT;
}
if (q->queue->pre_exec != NULL) {
*pre_exec = makeString(q->queue->pre_exec);
if (q->queue.pre_exec != NULL) {
*pre_exec = makeString(q->queue.pre_exec);
} else {
*pre_exec = NULL;
}
if (q->queue->post_exec != NULL) {
*post_exec = makeString(q->queue->post_exec);
if (q->queue.post_exec != NULL) {
*post_exec = makeString(q->queue.post_exec);
} else {
*post_exec = NULL;
}

16
src/controller/sched.c
View file

@ -341,7 +341,7 @@ static int sched_job_update_fun(int type, int situation, void* data, va_list ap)
xce_notify_data_t xdata = {
.event = NotifyData_data_schedUpdateJob,
.data = {
.sched_update_job = job->job
.sched_update_job = &(job->job)
}
};
/* clang-format on */
@ -419,11 +419,11 @@ static int sched_worker_change(int type, int situation, void* data,
int rc = 0;
xlist_t wlist;
TRACE("%s worker %s idx %d situation %d", __func__, w->host->name, w->idx,
TRACE("%s worker %s idx %d situation %d", __func__, w->host.name, w->idx,
situation);
collect_resval_with_idx(&(w->host->resvals), &(w->host->nresvals_to_sched),
&(w->host->resvals_to_sched));
collect_resval_with_idx(&(w->host.resvals), &(w->host.nresvals_to_sched),
&(w->host.resvals_to_sched));
mylock_lock(&(c->sched.lock));
if (c->sched.ep != NULL) {
@ -440,8 +440,8 @@ static int sched_worker_change(int type, int situation, void* data,
}
mylock_unlock(&(c->sched.lock));
free_resval_with_idx_list(&(w->host->resvals_to_sched),
w->host->nresvals_to_sched);
free_resval_with_idx_list(&(w->host.resvals_to_sched),
w->host.nresvals_to_sched);
return rc;
}
@ -458,7 +458,7 @@ static int sched_worker_join(int type, int situation, void* data, va_list ap)
return 0;
}
rc = key_find(c->workermgr, (void *)name, (void **)&w);
rc = key_find(c->workermgr, (void*)name, (void**)&w);
if (rc < 0) {
return rc;
}
@ -558,7 +558,7 @@ static int sched_queue_update(int type,
xce_notify_data_t xdata = {
.event = NotifyData_data_schedUpdateQueue,
.data = {
.sched_update_queue = q->queue
.sched_update_queue = &(q->queue)
}
};
/* clang-format on */

4
src/controller/task.c
View file

@ -26,7 +26,7 @@ static int run_tasks_on_host(cluster_t* c, endpoint_t* ep, uint64_t jobid,
return -ENOENT;
}
if ((w->ep == NULL) || (IS_WORKER_UNAVAIL(w->host->status))) {
if ((w->ep == NULL) || (IS_WORKER_UNAVAIL(w->host.status))) {
obj_deref(w);
return -EINVAL;
}
@ -59,7 +59,7 @@ static int check_task_hosts(cluster_t* c, xce_taskrun_t* runmsg)
return -ENOENT;
}
if ((w->ep == NULL) || IS_WORKER_UNAVAIL(w->host->status)) {
if ((w->ep == NULL) || IS_WORKER_UNAVAIL(w->host.status)) {
return -ENOENT;
}
}

90
src/controller/worker.c
View file

@ -70,7 +70,7 @@ static int update_val_res(cluster_t* c, worker_t* w, xce_resource_t* r,
array = &(c->sharedinst);
} else {
/* update host based resource */
array = &(w->host->resvals);
array = &(w->host.resvals);
}
if (idx < array->caps) {
@ -213,17 +213,16 @@ static int add_hostattrs_resvals(cluster_t* c, worker_t* w)
{
#define N_HOST_ATTRS (8)
xce_val_t values[N_HOST_ATTRS] = {
{ .which = Value_u_svalue, .u = { .svalue = w->host.attrs->hosttype } },
{ .which = Value_u_svalue,
.u = { .svalue = w->host->attrs->hosttype } },
{ .which = Value_u_svalue,
.u = { .svalue = w->host->attrs->hostmodel } },
.u = { .svalue = w->host.attrs->hostmodel } },
{ .which = Value_u_fvalue,
.u = { .fvalue = w->host->attrs->cpufactor } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host->attrs->maxcpus } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host->attrs->maxmem } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host->attrs->maxswap } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host->attrs->maxtmp } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host->attrs->ndisks } }
.u = { .fvalue = w->host.attrs->cpufactor } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host.attrs->maxcpus } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host.attrs->maxmem } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host.attrs->maxswap } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host.attrs->maxtmp } },
{ .which = Value_u_fvalue, .u = { .fvalue = w->host.attrs->ndisks } }
};
xce_resval_t resvals[N_HOST_ATTRS] = {
{ .name = "type", .val = &values[0] },
@ -266,7 +265,7 @@ int add_worker(void* cls, endpoint_t* ep, const char* host, hostattrs_t* attrs)
cluster_t* c = (cluster_t*)cls;
/* clang-format off */
worker_t worker = {
.host = &(xce_worker_t){
.host = (xce_worker_t){
.name = (char*)host,
}
};
@ -290,28 +289,28 @@ int add_worker(void* cls, endpoint_t* ep, const char* host, hostattrs_t* attrs)
return -errno;
}
worker.host->attrs = attrs;
worker.host.attrs = attrs;
w->host = makeWorker(worker.host);
if (w->host == NULL) {
w = makeWorker(c->workermgr, &(worker.host));
if (w == NULL) {
return -errno;
}
/* we should hold the update lock */
/* If admin makes the host in locked or closed, we need keep this */
w->host->status &= ~XCE_WORKER_UNAVAIL; /* status OK */
w->host.status &= ~XCE_WORKER_UNAVAIL; /* status OK */
w->ep = ep;
/* override host type? */
if (strcmp(hconfig->hosttype, "!") != 0) {
dereference(w->host->attrs->hosttype);
w->host->attrs->hosttype = makeString(hconfig->hosttype);
dereference(w->host.attrs->hosttype);
w->host.attrs->hosttype = makeString(hconfig->hosttype);
}
/* override host model */
if (strcmp(hconfig->hostmodel, "!") != 0) {
dereference(w->host->attrs->hostmodel);
w->host->attrs->hostmodel = makeString(hconfig->hostmodel);
dereference(w->host.attrs->hostmodel);
w->host.attrs->hostmodel = makeString(hconfig->hostmodel);
}
add_hostattrs_resvals(c, w);
idx = w->idx;
@ -319,31 +318,31 @@ int add_worker(void* cls, endpoint_t* ep, const char* host, hostattrs_t* attrs)
/* using auto detected or pre configured MXJ? */
nslots = hconfig->mxj;
if (nslots < 0) {
nslots = w->host->attrs->maxcpus;
nslots = w->host.attrs->maxcpus;
}
set_host_slots(c, w, nslots);
if (w->host->dispatch_window_config != NULL) {
free_timewindow_config(w->host->dispatch_window_config);
w->host->dispatch_window_config = NULL;
if (w->host.dispatch_window_config != NULL) {
free_timewindow_config(w->host.dispatch_window_config);
w->host.dispatch_window_config = NULL;
}
if (hconfig->dispatchwindow != NULL) {
w->host->dispatch_window_config
w->host.dispatch_window_config
= dup_timewindow_config(hconfig->dispatchwindow);
}
if (w->host->thresholds != NULL) {
free_resource_thresholds(w->host->thresholds, w->host->nthresholds);
w->host->nthresholds = 0;
w->host->thresholds = NULL;
if (w->host.thresholds != NULL) {
free_resource_thresholds(w->host.thresholds, w->host.nthresholds);
w->host.nthresholds = 0;
w->host.thresholds = NULL;
}
if (hconfig->thresholds != NULL) {
w->host->thresholds = dup_resource_thresholds(hconfig->thresholds,
hconfig->nthresholds);
w->host->nthresholds = hconfig->nthresholds;
w->host.thresholds = dup_resource_thresholds(hconfig->thresholds,
hconfig->nthresholds);
w->host.nthresholds = hconfig->nthresholds;
}
/* add bool resources to host */
@ -361,27 +360,18 @@ int restore_worker(void* cls, xce_worker_t host)
cluster_t* c = (cluster_t*)cls;
int idx = 0;
worker_t* w = NULL;
/* do work in add_worker() doing for normal worker join */
/* clang-format off */
worker_t worker = {
.host = &(xce_worker_t){
.name = (char*)host.name,
.status = host.status,
},
};
/* clang-format on */
w = calloc(1, sizeof(*w));
if (w == NULL) {
return -ENOMEM;
}
w->host = makeWorker(worker.host);
if (w->host == NULL) {
w = makeWorker(c->workermgr, &host);
if (w == NULL) {
return -errno;
}
add_hostattrs_resvals(c, w);
idx = w->idx;
w->host->status |= XCE_WORKER_UNAVAIL;
w->host.status |= XCE_WORKER_UNAVAIL;
DEREF_OBJ(w);
@ -476,7 +466,7 @@ int get_resval(cluster_t* c, char* resname, char* host, xce_val_t* val)
return -ENOENT;
}
array = &(w->host->resvals);
array = &(w->host.resvals);
} else {
array = &(c->sharedinst);
}
@ -497,7 +487,7 @@ int mark_worker_unavail(cluster_t* c, worker_t* w)
{
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_WORKER, HOOK_BEFORE_EVENT,
HOOK_DETAIL_WORKER_DEAD, c, w);
w->host->status |= XCE_WORKER_UNAVAIL;
w->host.status |= XCE_WORKER_UNAVAIL;
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_WORKER, HOOK_AFTER_EVENT,
HOOK_DETAIL_WORKER_DEAD, c, w);
@ -508,7 +498,7 @@ int mark_worker_back(cluster_t* c, worker_t* w)
{
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_WORKER, HOOK_BEFORE_EVENT,
HOOK_DETAIL_WORKER_UPDATE, c, w);
w->host->status &= ~XCE_WORKER_UNAVAIL;
w->host.status &= ~XCE_WORKER_UNAVAIL;
xce_run_hooks(&(c->hooksys), XCE_CONTROLLER_HOOK_WORKER, HOOK_AFTER_EVENT,
HOOK_DETAIL_WORKER_UPDATE, c, w);
@ -574,16 +564,16 @@ int worker_ctrl(cluster_t* c, xce_hostctrl_t* ctrl)
switch (ctrl->action) {
case HOST_OPEN:
if (!(w->host->status & XCE_WORKER_CLOSED)) {
if (!(w->host.status & XCE_WORKER_CLOSED)) {
goto clean;
}
w->host->status &= ~XCE_WORKER_CLOSED;
w->host.status &= ~XCE_WORKER_CLOSED;
break;
case HOST_CLOSE:
if (w->host->status & XCE_WORKER_CLOSED) {
if (w->host.status & XCE_WORKER_CLOSED) {
goto clean;
}
w->host->status |= XCE_WORKER_CLOSED;
w->host.status |= XCE_WORKER_CLOSED;
break;
default:
rc = -1;