webapps/term_extractor
18
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

'Priority que' logic implementation

Browse Source
This commit is contained in:
Kikimanox
2022-12-06 21:53:19 +01:00
parent 911bc8f0c9
commit d3cedd47b5
2 changed files with 143 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files
swagger_server/controllers/jobs_controller.py
+141 -64
View File
@@ -17,17 +17,39 @@ from werkzeug.datastructures import FileStorage
import threading import threading
import time import time
CLASSLA_CONCURANCE_LIMIT = 4 CLASSLA_SMALL_SIZE_LIMIT = 500 * 1e3 # 500 KB, aka.: 500 * 10^3
DOC2TEXT_CONCURANCE_LIMIT = 4
ATEAPI_CONCURANCE_LIMIT = 4 DOC2TEXT_SMALL_SIZE_LIMIT = 10 * 1e6 # 10 MB
ATEAPI_SMALL_SIZE_LIMIT = 2 * 1e6 # 2 MB
########################################
CLASSLA_CONCURANCE_LIMIT_BIG = 2
CLASSLA_CONCURANCE_LIMIT_SMALL = 2
DOC2TEXT_CONCURANCE_LIMIT_BIG = 2
DOC2TEXT_CONCURANCE_LIMIT_SMALL = 2
ATEAPI_CONCURANCE_LIMIT_BIG = 2
ATEAPI_CONCURANCE_LIMIT_SMALL = 2
IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT = 4 IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT = 4
########################################
classla_sem_big = threading.Semaphore(CLASSLA_CONCURANCE_LIMIT_BIG)
classla_sem_small = threading.Semaphore(CLASSLA_CONCURANCE_LIMIT_SMALL)
doc2text_sem_big = threading.Semaphore(DOC2TEXT_CONCURANCE_LIMIT_BIG)
doc2text_sem_small = threading.Semaphore(DOC2TEXT_CONCURANCE_LIMIT_SMALL)
ateapi_sem_big = threading.Semaphore(ATEAPI_CONCURANCE_LIMIT_BIG)
ateapi_sem_small = threading.Semaphore(ATEAPI_CONCURANCE_LIMIT_SMALL)
classla_sem = threading.Semaphore(CLASSLA_CONCURANCE_LIMIT)
doc2text_sem = threading.Semaphore(DOC2TEXT_CONCURANCE_LIMIT)
ateapi_sem = threading.Semaphore(ATEAPI_CONCURANCE_LIMIT)
izluscipoiskanju_sem = threading.Semaphore(IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT) izluscipoiskanju_sem = threading.Semaphore(IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT)
running_threads = {} running_threads = {} # <--- dict currently not used, was trying to figure out how to cancel workers mid execution,
# no luck with that yet
def delete_job(job_id): # noqa: E501 def delete_job(job_id): # noqa: E501
@@ -104,7 +126,19 @@ async def try_do_jobs():
ex.submit(try_do_jobs_izluscipoiskanju) ex.submit(try_do_jobs_izluscipoiskanju)
### Job looping # to pe je pod ex.submit
# sub = ex.submit(execute_ateapi_job, job)
# running_threads[job.id] = sub
# time.sleep(1)
# running_threads[job.id]
# preveri ce je done: ```running_threads[job.id].done()``` (vrne true false)
# was_canceled = running_threads[job.id].cancel()
# Todo: Mogoce kaksna druga opcija? Ampak verjetno ne, ne vidim (še?) kak prekicat ONGOING job
# To zgoraj preklice samo job, ki se se ni zacel, kar pa ni za ta use case uporabno.
### Picking jobs for looping
def try_do_jobs_izluscipoiskanju(): def try_do_jobs_izluscipoiskanju():
while True: while True:
try: try:
@@ -113,35 +147,37 @@ def try_do_jobs_izluscipoiskanju():
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 5) \ .where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 5) \
.limit(izluscipoiskanju_sem._value) .limit(izluscipoiskanju_sem._value)
with cf.ThreadPoolExecutor(max_workers=IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT) as ex: with cf.ThreadPoolExecutor(max_workers=IZLUSCI_PO_ISKANJU_CONCURANCE_LIMIT) as ex:
[ex.submit(execute_izluscipoiskanju_job, job) for job in unfinished_jobs] [ex.submit(execute_izluscipoiskanju_job, job, izluscipoiskanju_sem) for job in unfinished_jobs]
except Exception as e: except Exception as e:
print(f"Exception in try_do_jobs_ateapi") print(f"Exception in try_do_jobs_izluscipoiskanju")
traceback.print_exc() traceback.print_exc()
finally: finally:
time.sleep(3) time.sleep(3)
### Job looping ### Picking jobs for looping
def try_do_jobs_ateapi(): def try_do_jobs_ateapi():
while True: while True:
try: try:
if ateapi_sem._value > 0: unfinished_jobs_big = []
unfinished_jobs = Job.select() \ unfinished_jobs_small = []
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 4) \ if ateapi_sem_big._value > 0:
.limit(ateapi_sem._value) unfinished_jobs_big.extend(Job.select().where(Job.finished_on.is_null(), Job.started_on.is_null(),
with cf.ThreadPoolExecutor(max_workers=ATEAPI_CONCURANCE_LIMIT) as ex: Job.job_type == 4,
# [ex.submit(execute_ateapi_job, job) for job in unfinished_jobs] Job.input_size > ATEAPI_SMALL_SIZE_LIMIT) \
with cf.ThreadPoolExecutor(max_workers=ATEAPI_CONCURANCE_LIMIT) as ex: .limit(ateapi_sem_big._value))
for job in unfinished_jobs: if ateapi_sem_small._value > 0:
ex.submit(execute_ateapi_job, job) unfinished_jobs_small.extend(Job.select().where(Job.finished_on.is_null(), Job.started_on.is_null(),
# sub = ex.submit(execute_ateapi_job, job) Job.job_type == 4,
# running_threads[job.id] = sub Job.input_size <= ATEAPI_SMALL_SIZE_LIMIT) \
# time.sleep(1) .limit(ateapi_sem_small._value))
# running_threads[job.id]
# preveri ce je done: ```running_threads[job.id].done()``` (vrne true false)
# was_canceled = running_threads[job.id].cancel() if len(unfinished_jobs_big) > 0:
# Todo: Mogoce kaksna druga opcija? Ampak verjetno ne, ne vidim (še?) kak prekicat ONGOING job with cf.ThreadPoolExecutor(max_workers=ATEAPI_CONCURANCE_LIMIT_BIG) as ex:
[ex.submit(execute_ateapi_job, job, ateapi_sem_big) for job in unfinished_jobs_big]
if len(unfinished_jobs_small) > 0:
with cf.ThreadPoolExecutor(max_workers=ATEAPI_CONCURANCE_LIMIT_SMALL) as ex:
[ex.submit(execute_ateapi_job, job, ateapi_sem_small) for job in unfinished_jobs_small]
except Exception as e: except Exception as e:
print(f"Exception in try_do_jobs_ateapi") print(f"Exception in try_do_jobs_ateapi")
traceback.print_exc() traceback.print_exc()
@@ -149,27 +185,51 @@ def try_do_jobs_ateapi():
time.sleep(3) time.sleep(3)
### Job looping ### Picking jobs for looping
def try_do_jobs_classla(): def try_do_jobs_classla():
time.sleep(15) # wait for tokenizers to load for classla ... time.sleep(15) # wait for tokenizers to load for classla ...
while True: while True:
try: try:
if cl_utils.nlp_loaded: if not cl_utils.nlp_loaded:
if classla_sem._value > 0: raise Exception("NLP utils not loaded yet.")
unfinished_jobs_txt = Job.select() \
.where(Job.finished_on.is_null(), Job.job_type == 2,
Job.input_file.is_null(False)) \
.limit(classla_sem._value)
unfinished_jobs_no_txt = Job.select() \ unfinished_jobs_big = []
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 2, unfinished_jobs_small = []
Job.input_file.is_null()) \
.limit(classla_sem._value)
unfinished_jobs = [j for j in unfinished_jobs_txt] + [j for j in unfinished_jobs_no_txt] if classla_sem_big._value > 0:
unfinished_jobs = unfinished_jobs[:classla_sem._value] unfinished_jobs_txt = Job.select() \
with cf.ThreadPoolExecutor(max_workers=CLASSLA_CONCURANCE_LIMIT) as ex: .where(Job.finished_on.is_null(), Job.job_type == 2,
[ex.submit(execute_classla_job, job) for job in unfinished_jobs] Job.input_file.is_null(False), Job.input_size > CLASSLA_SMALL_SIZE_LIMIT) \
.limit(classla_sem_big._value)
unfinished_jobs_no_txt = Job.select() \
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 2,
Job.input_file.is_null(), Job.input_size > CLASSLA_SMALL_SIZE_LIMIT) \
.limit(classla_sem_big._value)
unfinished_jobs_big = [j for j in unfinished_jobs_txt] + [j for j in unfinished_jobs_no_txt]
unfinished_jobs_big = unfinished_jobs_big[:classla_sem_big._value]
if classla_sem_small._value > 0:
unfinished_jobs_txt = Job.select() \
.where(Job.finished_on.is_null(), Job.job_type == 2,
Job.input_file.is_null(False), Job.input_size <= CLASSLA_SMALL_SIZE_LIMIT) \
.limit(classla_sem_small._value)
unfinished_jobs_no_txt = Job.select() \
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type == 2,
Job.input_file.is_null(), Job.input_size <= CLASSLA_SMALL_SIZE_LIMIT) \
.limit(classla_sem_small._value)
unfinished_jobs_small = [j for j in unfinished_jobs_txt] + [j for j in unfinished_jobs_no_txt]
unfinished_jobs_small = unfinished_jobs_small[:classla_sem_small._value]
if len(unfinished_jobs_big) > 0:
with cf.ThreadPoolExecutor(max_workers=CLASSLA_CONCURANCE_LIMIT_BIG) as ex:
[ex.submit(execute_classla_job, job, doc2text_sem_big) for job in unfinished_jobs_big]
if len(unfinished_jobs_small) > 0:
with cf.ThreadPoolExecutor(max_workers=CLASSLA_CONCURANCE_LIMIT_SMALL) as ex:
[ex.submit(execute_classla_job, job, doc2text_sem_small) for job in unfinished_jobs_small]
except Exception as e: except Exception as e:
print(f"Exception in try_do_jobs_classla") print(f"Exception in try_do_jobs_classla")
@@ -178,16 +238,29 @@ def try_do_jobs_classla():
time.sleep(3) time.sleep(3)
### Job looping ### Picking jobs for looping
def try_do_jobs_doc2text(): def try_do_jobs_doc2text():
while True: while True:
try: try:
if doc2text_sem._value > 0: unfinished_jobs_big = []
unfinished_jobs = Job.select() \ unfinished_jobs_small = []
.where(Job.finished_on.is_null(), Job.started_on.is_null(), Job.job_type << [1, 12, 3, 32]) \ if doc2text_sem_big._value > 0:
.limit(doc2text_sem._value) unfinished_jobs_big.extend(Job.select().where(Job.finished_on.is_null(), Job.started_on.is_null(),
with cf.ThreadPoolExecutor(max_workers=DOC2TEXT_CONCURANCE_LIMIT) as ex: Job.job_type << [1, 12, 3, 32],
[ex.submit(execute_doc2text_job, job) for job in unfinished_jobs] Job.input_size > DOC2TEXT_SMALL_SIZE_LIMIT) \
.limit(doc2text_sem_big._value))
if doc2text_sem_small._value > 0:
unfinished_jobs_small.extend(Job.select().where(Job.finished_on.is_null(), Job.started_on.is_null(),
Job.job_type << [1, 12, 3, 32],
Job.input_size <= DOC2TEXT_SMALL_SIZE_LIMIT) \
.limit(doc2text_sem_small._value))
if len(unfinished_jobs_big) > 0:
with cf.ThreadPoolExecutor(max_workers=DOC2TEXT_CONCURANCE_LIMIT_BIG) as ex:
[ex.submit(execute_doc2text_job, job, doc2text_sem_big) for job in unfinished_jobs_big]
if len(unfinished_jobs_small) > 0:
with cf.ThreadPoolExecutor(max_workers=DOC2TEXT_CONCURANCE_LIMIT_SMALL) as ex:
[ex.submit(execute_doc2text_job, job, doc2text_sem_small) for job in unfinished_jobs_small]
except Exception as e: except Exception as e:
print(f"Exception in try_do_jobs_doc2text") print(f"Exception in try_do_jobs_doc2text")
traceback.print_exc() traceback.print_exc()
@@ -195,13 +268,14 @@ def try_do_jobs_doc2text():
time.sleep(3) time.sleep(3)
async def prep_jobs(tasks): # async def prep_jobs(tasks):
await asyncio.gather(*tasks) # await asyncio.gather(*tasks)
def execute_doc2text_job(job: Job): ####### JOB EXECUTION LOGIC
def execute_doc2text_job(job: Job, sem: threading.Semaphore):
try: try:
doc2text_sem.acquire() sem.acquire()
del_file = False del_file = False
job.started_on = datetime.datetime.utcnow() job.started_on = datetime.datetime.utcnow()
job.save() job.save()
@@ -209,7 +283,7 @@ def execute_doc2text_job(job: Job):
tmp_file_path = job.input_file tmp_file_path = job.input_file
if not os.path.exists(tmp_file_path): if not os.path.exists(tmp_file_path):
job.finished_on = datetime.datetime.utcnow() job.finished_on = datetime.datetime.utcnow()
job.job_output = "ERROR - Temporary file went missing, couldn't properly finish job" job.job_output = "ERROR - Temporary file went missing, couldn't properly finish job. Please try executing the job again."
job.save() job.save()
return return
@@ -241,12 +315,13 @@ def execute_doc2text_job(job: Job):
job.started_on = None job.started_on = None
job.save() job.save()
finally: finally:
doc2text_sem.release() sem.release()
def execute_classla_job(job: Job): ####### JOB EXECUTION LOGIC
def execute_classla_job(job: Job, sem: threading.Semaphore):
try: try:
classla_sem.acquire() sem.acquire()
job.started_on = datetime.datetime.utcnow() job.started_on = datetime.datetime.utcnow()
job.save() job.save()
conllu, status = cl_utils.raw_text_to_conllu(job.job_input) conllu, status = cl_utils.raw_text_to_conllu(job.job_input)
@@ -261,12 +336,13 @@ def execute_classla_job(job: Job):
job.save() job.save()
print(f"Unexpected error at job {job.id}") print(f"Unexpected error at job {job.id}")
finally: finally:
classla_sem.release() sem.release()
def execute_ateapi_job(job: Job): ####### JOB EXECUTION LOGIC
def execute_ateapi_job(job: Job, sem: threading.Semaphore):
try: try:
ateapi_sem.acquire() sem.acquire()
job.started_on = datetime.datetime.utcnow() job.started_on = datetime.datetime.utcnow()
job.save() job.save()
info = json.loads(job.job_input) info = json.loads(job.job_input)
@@ -296,12 +372,13 @@ def execute_ateapi_job(job: Job):
job.save() job.save()
print(f"Unexpected error at job {job.id}") print(f"Unexpected error at job {job.id}")
finally: finally:
ateapi_sem.release() sem.release()
def execute_izluscipoiskanju_job(job: Job): ####### JOB EXECUTION LOGIC
def execute_izluscipoiskanju_job(job: Job, sem: threading.Semaphore):
try: try:
izluscipoiskanju_sem.acquire() sem.acquire()
job.started_on = datetime.datetime.utcnow() job.started_on = datetime.datetime.utcnow()
job.save() job.save()
info = json.loads(job.job_input) info = json.loads(job.job_input)
@@ -316,7 +393,7 @@ def execute_izluscipoiskanju_job(job: Job):
job.save() job.save()
print(f"Unexpected error at job {job.id}") print(f"Unexpected error at job {job.id}")
finally: finally:
izluscipoiskanju_sem.release() sem.release()
clear_up_unfinished_jobs() clear_up_unfinished_jobs()
swagger_server/requets_db/models/vrsta.py
+2 -1
View File
@@ -65,7 +65,8 @@ class JobManager:
pathlib.Path('tmp').mkdir(exist_ok=True) pathlib.Path('tmp').mkdir(exist_ok=True)
job_input: werkzeug.datastructures.FileStorage job_input: werkzeug.datastructures.FileStorage
job_input.save(tmp_file) job_input.save(tmp_file)
job, is_new = Job.get_or_create(job_type=job_type, input_file=tmp_file, input_size=-1) job, is_new = Job.get_or_create(job_type=job_type, input_file=tmp_file,
input_size=os.stat(tmp_file).st_size)
return job, is_new return job, is_new
except Exception as e: except Exception as e: