基于以前的一篇文章,celery+Rabbit MQ的安裝和使用,
本文更加詳細的介紹如何安裝和使用celey, Rabbit MQ。
并記錄在使用celery時遇到的一些問題。
1.安裝 Rabbit MQ
在 OS X上,直接執行如下指令:
$ brew install rabbitmq
其他作業系統下的安裝可以參考安裝 RabbitMQ
啟動RabbitMQ
$ sudo rabbitmq-server
你也可以添加 -detached 屬性來讓它在背景運作(注意:隻有一個破折号):
$ sudo rabbitmq-server -detached
檢視RabbitMQ狀态
$ sudo rabbitmqctl status
停止RabbitMQ
永遠不要用 kill 停止 RabbitMQ 伺服器,而是應該用 rabbitmqctl 指令:
$ sudo rabbitmqctl stop
添加使用者
預設使用者guest,密碼guest,隻允許本地通路,如需遠端通路,需要設定.
$ sudo rabbitmqctl add_user test 123456
Adding user "test" ...
添加虛拟主機,并賦予使用者test權限
$ sudo rabbitmqctl add_vhost myvhost
Adding vhost "myvhost" ...
$ sudo rabbitmqctl set_permissions -p myvhost test ".*" ".*" ".*"
Setting permissions for user "test" in vhost "myvhost" ...
2.安裝celery
2.1 建立虛拟環境,并安裝celery
$ mkdir celery_demo
$ cd celery_demo
$ virtualenv -p python3 venv3
$ ./venv3/bin/pip install celery
2.2 配置celery
建立配置檔案 celeryconfig.py,裡面包含
BROKER_URL
、
CELERYD_LOG_FORMAT
、
CELERY_ROUTES
.
# celeryconfig.py
RABBIT_MQ = {
'HOST': '127.0.0.1',
'PORT': 5672,
'USER': 'test',
'PASSWORD': '123456'
}
# broker
BROKER_URL = 'amqp://%s:%s@%s:%s/myvhost' % (RABBIT_MQ['USER'], RABBIT_MQ['PASSWORD'], RABBIT_MQ['HOST'], RABBIT_MQ['PORT'])
# celery日志格式
CELERYD_LOG_FORMAT = '[%(asctime)s] [%(levelname)s] %(message)s'
CELERY_ROUTES = {
'demo_task.add': {'queue': 'sunday'},
}
其中,參數定義如下:
-
指定了broker資訊,即消息隊列的位址。BROKER_URL
-
指定了日志格式。CELERYD_LOG_FORMAT
-
指定了路由資訊,即調用CELERY_ROUTES
後,消息具體放入哪個隊列,這裡是隊列名稱為demo_task.add
。sunday
2.3 啟動消費者
消費者
消費者代碼如下:
# demo_task.py
from celery import Celery
app = Celery("orange", backend='amqp')
app.config_from_object("celeryconfig")
@app.task
def add(x, y):
return x + y
首先,建立
Celery
執行個體,從檔案中讀取配置。
接着,定義task。
其中,在建立
Celery
執行個體時,參數
backend
指定了結果存儲後端,用于追蹤task執行狀态和結果。這裡使用
amqp
,即使用RabbitMQ儲存結果。預設情況下,
backend
參數是關閉的。
使用
celery worker
啟動消費者
./venv3/bin/celery worker -A demo_task -Q sunday --loglevel=info -f app.log
[email protected] v4.3.0 (rhubarb)
Darwin-18.2.0-x86_64-i386-64bit 2019-04-20 17:49:40
[config]
.> app: orange:0x109ca0400
.> transport: amqp://test:**@127.0.0.1:5672/myvhost
.> results: disabled://
.> concurrency: 8 (prefork)
.> task events: OFF (enable -E to monitor tasks in this worker)
[queues]
.> sunday exchange=sunday(direct) key=sunday
[tasks]
. demo_task.add
其中,參數定義如下:
- 參數
是app name,即定義celery的檔案。-A
- 參數
指定了隊列的名稱,如果不指定,預設為-Q
。celery
- 參數
指定了日志列印檔案。-f
可以通過以下指令檢視更多幫助資訊:
-
檢視celery的選項celery help
-
檢視worker的選項celery worker --help
2.4 啟動生産者
生産者
代碼如下api.py所示:
# api.py
from demo_task import add
print("start...")
result = add.apply_async((1, 2))
print("result:", result)
print(result.ready())
print("end...")
代碼中,将
1
、
2
參數放入消息隊列。
celery配置
backend
參數後, 調用任務時,會傳回
AsyncResult
執行個體。
AsyncResult
的
ready()
方法可以檢視任務是否完成處理。
執行api.py
執行結果分兩種情況:worker啟動和沒有啟動。
當celery worker啟動的時候,結果如下:
./venv3/bin/python api.py
start...
result: 60dd0ab6-5fa2-4190-954e-584eb519384f
True
end...
可以看到,task很快得到執行,結果狀态為True。
當celery worker沒有啟動的時候,結果如下:
./venv3/bin/python api.py
start...
result: 7280466f-73cd-44ec-85e7-5ad4f079a797
False
end...
可以看到,結果狀态為False。
直接擷取結果
可以使用
get()
函數等待任務完成,但這很少使用,因為它把異步調用變成了同步調用。
對于
get()
的使用,按照是否設定逾時參數,分為兩種:不使用逾時參數timeout和使用逾時參數timeout。
(1)不使用逾時參數timeout
預設timeout為
None
,不會逾時,而是阻塞住。
通過tcpdump抓包,可以發現,
get
調用後,每隔30s,RabbitMQ(server端)向生産者API(client端)發送心跳包,共發了5次。
最後,過了3分鐘,RabbitMQ(server端)關閉連接配接,生産者API(client端)也關閉連接配接,報錯:
start...
result: aae4fb20-823c-4674-8bff-28429f14d5d7
False
Traceback (most recent call last):
File "api.py", line 13, in <module>
result.get()
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/result.py", line 226, in get
on_message=on_message,
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/base.py", line 496, in wait_for_pending
no_ack=no_ack,
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 146, in wait_for
on_interval=on_interval)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 223, in consume
conn, consumer, timeout, on_interval)[task_id]
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 204, in drain_events
wait(timeout=1)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/kombu/connection.py", line 315, in drain_events
return self.transport.drain_events(self.connection, **kwargs)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/kombu/transport/pyamqp.py", line 103, in drain_events
return connection.drain_events(**kwargs)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/amqp/connection.py", line 500, in drain_events
while not self.blocking_read(timeout):
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/amqp/connection.py", line 505, in blocking_read
frame = self.transport.read_frame()
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/amqp/transport.py", line 256, in read_frame
frame_header = read(7, True)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/amqp/transport.py", line 448, in _read
raise IOError('Server unexpectedly closed connection')
OSError: Server unexpectedly closed connection
(2)使用逾時參數timeout
如果使用
timeout
參數,
get(timeout)
調用發起後, 當超過
timeout
指定的時間仍然沒有獲得結果,會逾時報錯。
例如,調用
get(timeout=1)
:
... ...
# sync
print(result.get(timeout=1))
print("end...")
如果worker沒有啟動或者worker處理逾時,
get
會報逾時錯誤:
start...
result: 190ad871-aae2-48ac-a18e-962cd2d537b7
False
Traceback (most recent call last):
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 146, in wait_for
on_interval=on_interval)
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 223, in consume
conn, consumer, timeout, on_interval)[task_id]
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 202, in drain_events
raise socket.timeout()
socket.timeout
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "api.py", line 14, in <module>
print(result.get(timeout=1))
File "/workspace//celery_demo/venv3/lib/python3.6/site-packages/celery/result.py", line 226, in get
on_message=on_message,
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/base.py", line 496, in wait_for_pending
no_ack=no_ack,
File "/workspace/celery_demo/venv3/lib/python3.6/site-packages/celery/backends/amqp.py", line 148, in wait_for
raise TimeoutError('The operation timed out.')
celery.exceptions.TimeoutError: The operation timed out.
2.5 檢視結果
檢視app.log中的日志
[2019-04-20 20:53:59,327] [INFO] Connected to amqp://test:**@127.0.0.1:5672/myvhost
[2019-04-20 20:53:59,345] [INFO] mingle: searching for neighbors
[2019-04-20 20:54:00,379] [INFO] mingle: all alone
[2019-04-20 20:54:00,441] [INFO] [email protected] ready.
[2019-04-20 20:55:46,498] [INFO] Received task: demo_task.add[1c8d47bd-449d-4bd9-b4db-819777081d23]
[2019-04-20 20:55:46,561] [INFO] Task demo_task.add[1c8d47bd-449d-4bd9-b4db-819777081d23] succeeded in 0.060540573904290795s: 3
3.任務過期
調用
apply_async
時使用參數
expires
,則表示任務有逾時時間,超過這個時間後,task不會得到執行。
result = add.apply_async((1, 2), expires=10)
當worker沒有啟動或者其他異常情況下,會出現任務逾時,不被執行。
檢視日志,可以看到task過期,不會被執行。
[2019-04-20 21:32:45,828] [INFO] Received task: demo_task.add[12691c27-3f2f-4c96-9b4e-54636e20d0eb] expires:[2019-04-20 13:31:51.030013+00:00]
[2019-04-20 21:32:45,829] [INFO] Discarding revoked task: demo_task.add[12691c27-3f2f-4c96-9b4e-54636e20d0eb]
3.參考
Celery 初步
calling API
Configuration and defaults
expiration