I'm just getting into Celery chains in my Django project. I have the following function:
def orchestrate_tasks_for_account(account_id):
# Get the account, set status to 'SYNC' until the chain is complete
account = Account.objects.get(id=account_id)
account.status = "SYNC"
account.save()
chain = task1.s(account_id) | task2.s() | task3.s()
chain()
# if any of the tasks in the chain failed, set account.status = 'ERROR'
# else set the account.status = 'OK'
The chain works as expected, but I'm not sure how to take feedback from the chain and update the account based on the results
In other words, I'd like to set the account status to 'ERROR' if any of the tasks in the chain fail, otherwise I'd like to set the account status to 'OK'
I'm confused by the Celery documentation on how to handle an error with an if/else like I've commented in the last two lines above.
Does anyone have experience with this?
Ok - here's what I came up with
I've leveraged the waiting library in this solution
from celery import chain
from waiting import wait
def orchestrate_tasks_for_account(account_id):
account = Account.objects.get(id=account_id)
account.status = "SYNC"
account.save()
job = chain(
task1.s(account_id),
task2.s(),
task3.s()
)
result = job.apply_async()
wait(
lambda: result.ready(), # when async job is completed...
timeout_seconds=1800, # wait 1800 seconds (30 minutes)
waiting_for="task orchestration to complete"
)
if result.successful():
account.status = 'OK'
else:
account.status = 'ERROR'
account.save()
I am open to suggestions to make this better!
Related
I am trying to add a procedure to pop-up a modal dialog inside a plug-in.
Its purpose is to query a response at designated steps within the control-flow of the plug-in (not just acquire parameters at its start).
I have tried using gtk - I get a dialog but it is asynchronous - the plugin continues execution. It needs to operate as a synchronous function.
I have tried registering a plugin in order to take advantage of the gimpfu start-up dialogue for same. By itself, it works; it shows up in the procedural db when queried. But I never seem to be able to actually invoke it from within another plug-in - its either an execution error or wrong number of arguments no matter how many permutations I try.
[Reason behind all of this nonsense: I have written a lot of extension Python scripts for PaintShopPro. I have written a App package (with App.Do, App.Constants, Environment and the like that lets me begin to port those scripts to GIMP -- yes it is perverse, and yes sometimes the code just has to be rewritten, but for a lot of what I actual use in the PSP.API it is sufficient.
However, debugging and writing the module rhymes with witch. So. I am trying to add emulation of psp's "SetExecutionMode" (ie interactive). If
set, the intended behavior is that the App.Do() method will "pause" after/before it runs the applicable psp emulation code by popping up a simple message dialog.]
A simple modal dialogue within a gimp python-fu plug-in can be implemented via gtk's Dialog interface, specifically gtk.MessageDialog.
A generic dialog can be created via
queryDialogue = gtk.MessageDialog(None, gtk.DIALOG_DESTROY_WITH_PARENT \
gtk.MESSAGE_QUESTION, \
gtk.BUTTONS_OK_CANCEL, "")
Once the dialog has been shown,
a synchronous response may be obtained from it
queryDialogue.show()
response = queryDialogue.run()
queryDialogue.hide()
The above assumes that the dialog is not created and thence destroyed after each use.
In the use case (mentioned in the question) of a modal dialog to manage single stepping through a pspScript in gimp via an App emulator package, the dialogue message contents need to be customized for each use. [Hence, the "" for the message argument in the Constructor. [more below]]
In addition, the emulator must be able to accept a [cancel] response to 'get out of Dodge' - ie quit the entire plug-in (gracefully). I could not find a gimpfu interface for the latter, (and do not want to kill the app entirely via gimp.exit()). Hence, this is accomplished by raising a custom Exception class [appTerminate] within the App pkg and catching the exception in the outer-most scope of the plugin. When caught, then, the plug-in returns (exits).[App.Do() can not return a value to indicate continue/exit/etc, because the pspScripts are to be included verbatim.]
The following is an abbreviated skeleton of the solution -
a plug-in incorporating (in part) a pspScript
the App.py pkg supplying the environment and App.Do() to support the pspScript
a Map.py pkg supporting how pspScripts use dot-notation for parameters
App.py demonstrates creation, customization and use of a modal dialog - App.doContinue() displays the dialogue illustrating how it can be customized on each use.
App._parse() parses the pspScript (excerpt showing how it determines to start/stop single-step via the dialogue)
App._exec() implements the pspScript commands (excerpt showing how it creates the dialogue, identifies the message widget for later customization, and starts/stops its use)
# App.py (abbreviated)
#
import gimp
import gtk
import Map # see https://stackoverflow.com/questions/2352181/how-to- use-a-dot-to-access-members-of-dictionary
from Map import *
pdb = gimp.pdb
isDialogueAvailable = False
queryDialogue = None
queryMessage = None
Environment = Map({'executionMode' : 1 })
_AutoActionMode = Map({'Match' : 0})
_ExecutionMode = Map({'Default' : 0}, Silent=1, Interactive=2)
Constants = Map({'AutoActionMode' : _AutoActionMode}, ExecutionMode=_ExecutionMode ) # etc...
class appTerminate(Exception): pass
def Do(eNvironment, procedureName, options = {}):
global appTerminate
img = gimp.image_list()[0]
lyr = pdb.gimp_image_get_active_layer(img)
parsed = _parse(img, lyr, procedureName, options)
if eNvironment.executionMode == Constants.ExecutionMode.Interactive:
resp = doContinue(procedureName, parsed.detail)
if resp == -5: # OK
print procedureName # log to stdout
if parsed.valid:
if parsed.isvalid:
_exec(img, lyr, procedureName, options, parsed, eNvironment)
else:
print "invalid args"
else:
print "invalid procedure"
elif resp == -6: # CANCEL
raise appTerminate, "script cancelled"
pass # terminate plugin
else:
print procedureName + " skipped"
pass # skip execution, continue
else:
_exec(img, lyr, procedureName, options, parsed, eNvironment)
return
def doContinue(procedureName, details):
global queryMessage, querySkip, queryDialogue
# - customize the dialog -
if details == "":
msg = "About to execute procedure \n "+procedureName+ "\n\nContinue?"
else:
msg = "About to execute procedure \n "+procedureName+ "\n\nDetails - \n" + details +"\n\nContinue?"
queryMessage.set_text(msg)
queryDialogue.show()
resp = queryDialogue.run() # get modal response
queryDialogue.hide()
return resp
def _parse(img, lyr, procedureName, options):
# validate and interpret App.Do options' semantics vz gimp
if procedureName == "Selection":
isValid=True
# ...
# parsed = Map({'valid' : True}, isvalid=True, start=Start, width=Width, height=Height, channelOP=ChannelOP ...
# /Selection
# ...
elif procedureName == "SetExecutionMode":
generalOptions = options['GeneralSettings']
newMode = generalOptions['ExecutionMode']
if newMode == Constants.ExecutionMode.Interactive:
msg = "set mode interactive/single-step"
else:
msg = "set mode silent/run"
parsed = Map({'valid' : True}, isvalid=True, detail=msg, mode=newMode)
# /SetExecutionMode
else:
parsed = Map({'valid' : False})
return parsed
def _exec(img, lyr, procedureName, options, o, eNvironment):
global isDialogueAvailable, queryMessage, queryDialogue
#
try:
# -------------------------------------------------------------------------------------------------------------------
if procedureName == "Selection":
# pdb.gimp_rect_select(img, o.start[0], o.start[1], o.width, o.height, o.channelOP, ...
# /Selection
# ...
elif procedureName == "SetExecutionMode":
generalOptions = options['GeneralSettings']
eNvironment.executionMode = generalOptions['ExecutionMode']
if eNvironment.executionMode == Constants.ExecutionMode.Interactive:
if isDialogueAvailable:
queryDialogue.destroy() # then clean-up and refresh
isDialogueAvailable = True
queryDialogue = gtk.MessageDialog(None, gtk.DIALOG_DESTROY_WITH_PARENT, gtk.MESSAGE_QUESTION, gtk.BUTTONS_OK_CANCEL, "")
queryDialogue.set_title("psp/APP.Do Emulator")
queryDialogue.set_size_request(450, 180)
aqdContent = queryDialogue.children()[0]
aqdHeader = aqdContent.children()[0]
aqdMsgBox = aqdHeader.children()[1]
aqdMessage = aqdMsgBox.children()[0]
queryMessage = aqdMessage
else:
if isDialogueAvailable:
queryDialogue.destroy()
isDialogueAvailable = False
# /SetExecutionMode
else: # should not get here (should have been screened by parse)
raise AssertionError, "unimplemented PSP procedure: " + procedureName
except:
raise AssertionError, "App.Do("+procedureName+") generated an exception:\n" + sys.exc_info()
return
A skeleton of the plug-in itself. This illustrates incorporating a pspScript which includes a request for single-step/interactive execution mode, and thus the dialogues. It catches the terminate exception raised via the dialogue, and then terminates.
def generateWebImageSet(dasImage, dasLayer, title, mode):
try:
img = dasImage.duplicate()
# ...
bkg = img.layers[-1]
frameWidth = 52
start = bkg.offsets
end = (start[0]+bkg.width, start[1]+frameWidth)
# pspScript: (snippet included verbatim)
# SetExecutionMode / begin interactive single-step through pspScript
App.Do( Environment, 'SetExecutionMode', {
'GeneralSettings': {
'ExecutionMode': App.Constants.ExecutionMode.Interactive
}
})
# Selection
App.Do( Environment, 'Selection', {
'General' : {
'Mode' : 'Replace',
'Antialias' : False,
'Feather' : 0
},
'Start': start,
'End': end
})
# Promote
App.Do( Environment, 'SelectPromote' )
# und_so_weiter ...
except App.appTerminate:
raise AssertionError, "script cancelled"
# /generateWebImageSet
# _generateFloatingCanvasSetWeb.register -----------------------------------------
#
def generateFloatingCanvasSetWeb(dasImage, dasLayer, title):
mode="FCSW"
generateWebImageSet(dasImage, dasLayer, title, mode)
register(
"generateFloatingCanvasSetWeb",
"Generate Floating- Frame GW Canvas Image Set for Web Page",
"Generate Floating- Frame GW Canvas Image Set for Web Page",
"C G",
"C G",
"2019",
"<Image>/Image/Generate Web Imagesets/Floating-Frame Gallery-Wrapped Canvas Imageset...",
"*",
[
( PF_STRING, "title", "title", "")
],
[],
generateFloatingCanvasSetWeb)
main()
I realize that this may seem like a lot of work just to be able to include some pspScripts in a gimp plug-in, and to be able to single-step through the emulation. But we are talking about maybe 10K lines of scripts (and multiple scripts).
However, if any of this helps anyone else with dialogues inside plug-ins, etc., so much the better.
Our airflow implementation sends out http requests to get services to do tasks. We want those services to let airflow know when they complete their task, so we are sending a callback url to the service which they will call when their task is complete. I can't seem to find a callback sensor, however. How do people handle this normally?
There is no such thing as a callback or webhook sensor in Airflow. The sensor definition follows as taken from the documentation:
Sensors are a certain type of operator that will keep running until a certain criterion is met. Examples include a specific file landing in HDFS or S3, a partition appearing in Hive, or a specific time of the day. Sensors are derived from BaseSensorOperator and run a poke method at a specified poke_interval until it returns True.
This means that a sensor is an operator that performs polling behavior on external systems. In that sense, your external services should have a way of keeping state for each executed task - either internally or externally - so that a polling sensor can check on that state.
This way you can use for example the airflow.operators.HttpSensor that polls an HTTP endpoint until a condition is met. Or even better, write your own custom sensor that gives you the opportunity to do more complex processing and keep state.
Otherwise, if the service outputs data in a storage system you can use a sensor that polls a database for example. I believe you get the idea.
I'm attaching a custom operator example that I've written for integrating with the Apache Livy API. The sensor does two things: a) submits a Spark job through the REST API and b) waits for the job to be completed.
The operator extends the SimpleHttpOperator and at the same time implements the HttpSensor thus combining both functionalities.
class LivyBatchOperator(SimpleHttpOperator):
"""
Submits a new Spark batch job through
the Apache Livy REST API.
"""
template_fields = ('args',)
ui_color = '#f4a460'
#apply_defaults
def __init__(self,
name,
className,
file,
executorMemory='1g',
driverMemory='512m',
driverCores=1,
executorCores=1,
numExecutors=1,
args=[],
conf={},
timeout=120,
http_conn_id='apache_livy',
*arguments, **kwargs):
"""
If xcom_push is True, response of an HTTP request will also
be pushed to an XCom.
"""
super(LivyBatchOperator, self).__init__(
endpoint='batches', *arguments, **kwargs)
self.http_conn_id = http_conn_id
self.method = 'POST'
self.endpoint = 'batches'
self.name = name
self.className = className
self.file = file
self.executorMemory = executorMemory
self.driverMemory = driverMemory
self.driverCores = driverCores
self.executorCores = executorCores
self.numExecutors = numExecutors
self.args = args
self.conf = conf
self.timeout = timeout
self.poke_interval = 10
def execute(self, context):
"""
Executes the task
"""
payload = {
"name": self.name,
"className": self.className,
"executorMemory": self.executorMemory,
"driverMemory": self.driverMemory,
"driverCores": self.driverCores,
"executorCores": self.executorCores,
"numExecutors": self.numExecutors,
"file": self.file,
"args": self.args,
"conf": self.conf
}
print payload
headers = {
'X-Requested-By': 'airflow',
'Content-Type': 'application/json'
}
http = HttpHook(self.method, http_conn_id=self.http_conn_id)
self.log.info("Submitting batch through Apache Livy API")
response = http.run(self.endpoint,
json.dumps(payload),
headers,
self.extra_options)
# parse the JSON response
obj = json.loads(response.content)
# get the new batch Id
self.batch_id = obj['id']
log.info('Batch successfully submitted with Id %s', self.batch_id)
# start polling the batch status
started_at = datetime.utcnow()
while not self.poke(context):
if (datetime.utcnow() - started_at).total_seconds() > self.timeout:
raise AirflowSensorTimeout('Snap. Time is OUT.')
sleep(self.poke_interval)
self.log.info("Batch %s has finished", self.batch_id)
def poke(self, context):
'''
Function that the sensors defined while deriving this class should
override.
'''
http = HttpHook(method='GET', http_conn_id=self.http_conn_id)
self.log.info("Calling Apache Livy API to get batch status")
# call the API endpoint
endpoint = 'batches/' + str(self.batch_id)
response = http.run(endpoint)
# parse the JSON response
obj = json.loads(response.content)
# get the current state of the batch
state = obj['state']
# check the batch state
if (state == 'starting') or (state == 'running'):
# if state is 'starting' or 'running'
# signal a new polling cycle
self.log.info('Batch %s has not finished yet (%s)',
self.batch_id, state)
return False
elif state == 'success':
# if state is 'success' exit
return True
else:
# for all other states
# raise an exception and
# terminate the task
raise AirflowException(
'Batch ' + str(self.batch_id) + ' failed (' + state + ')')
Hope this will help you a bit.
I have a case where I am able to successfully trigger a build in buildbot by using sendchange command. While this works, I am unable to find a command to check if the build that was triggered by sendchange has finished.
Is there a way to achieve this in buildbot?
Thanks!
Since buildbot is asynchronous, you will need to poll the builder for builds that match your sendchange, and then poll that build for build status. Using e.g. python, it's fairly trivial using requests (https://pypi.python.org/pypi/requests) to retrieve a build's json and examine the state from the command line.
The "API" in this case is to use requests.get(url).json() and traverse the buildbot builds looking for your change request. The buildbot json is documented in the "REST API" section of the docs (http://docs.buildbot.net/latest/developer/rest.html), you'll have to hunt to figure out how change requests are stored.
Here's some code that will get you started:
import pprint, requests
def get_url_base(serv,port):
return 'http://%(serv)s:%(port)d' % locals()
def get_bldr_json(serv,port,bldr):
url = 'http://%(serv)s:%(port)d/json/builders/%(bldr)s' % locals()
print "get_bldr_json: %s ..." % url
jdata = requests.get(url).json()
print "DEBUG: get_bldr_json:", pprint.pformat(jdata)
return jdata
def get_bld_json(serv,port,bldr,bnum):
url = 'http://%(serv)s:%(port)d/json/builders/%(bldr)s/builds/%(bnum)s' % locals()
print "get_bld_json: %s ..." % url
jdata = requests.get(url).json()
print "DEBUG: get_bld_json:", pprint.pformat(jdata)
return jdata
# you'll have to set these values for your buildbot
serv, port, bldr = ('hexbotserver', 8010, 'buildername')
jdata = get_bldr_json(serv,port,bldr)
for bnum in jdata['cachedBuilds']:
jdata = get_bld_json(serv,port,bldr,bnum)
print "build properties:"
pprint.pprint(dict(jdata)['properties'])
I have two tasks. the task "heavy_task" need a concurrency of 1 and the "lite_task" need a concurrency of 4
#task
def lite_task():
tabla = Proc_Carga()
sp = tabla.carga()
return None
#task()
def heavy_task(idprov,pfecha):
conci = Buscar_Conci()
spconc = conci.buscarcon(idprov,pfecha)
return None
I define the routes in my settings.py file:
BROKER_URL = 'redis://localhost:6379/0'
CELERY_IMPORTS = ("pc.tasks", )
CELERY_ACCEPT_CONTENT = ['json']
CELERY_TASK_SERIALIZER = 'json'
CELERY_RESULT_SERIALIZER = 'json'
CELERY_RESULT_BACKEND='djcelery.backends.cache:CacheBackend'
CELERY_ROUTES = {"tasks.heavy_task": {"queue": "heavy"},"tasks.lite_task": {"queue": "lite"}}
I try to excute two workers specifying the concurrency in this way
celery multi start heavy lite -A provcon -c:heavy 1 -c:lite 3
When first call the tasks heavy_task work fine and the concurrency works,
but after call the task lite_task the concurency for the queue heavy change.
I try this:
celery -A provcon worker -Q heavy -c 1
And when I execute the task heavy_task, the routing won't work and the task is not executed.
but if use this:
celery -A provcon -c 1
everything works fine, but I can only execute one task at time, and I need to be able to execute the heavy_task with a concurrency of 1 and the lite_task with a concurrency of 3
Any advice
I try different settings to make the queues works and finally I did it.
In the tasks.py file I harcoded the the queue in the task decorator
#task(queue = 'heavy')
To run the workers I use this:
celery multi start lite_w heavy_w -A provcon -Q:heavy_w heavy -Q:carga_w lite -l info -c:heavy_w 1 -c:lite_w 3 -E
And i delete the routing settings from the settings.py file. My settings are this:
BROKER_URL = 'redis://localhost:6379/0'
CELERY_IMPORTS = ("pc.tasks", )
CELERY_ACCEPT_CONTENT = ['json']
CELERY_TASK_SERIALIZER = 'json'
CELERY_RESULT_SERIALIZER = 'json'
CELERY_RESULT_BACKEND='djcelery.backends.cache:CacheBackend'
I'm using resque/redis on heroku to send emails as a background job. It works fine for the first email I send, but after that I get the error: (ActiveRecord::StatementInvalid: PG::Error: SSL error: decryption failed or bad record mac :) ...
I've seen the other questions/answers that say to add to an initializer:
Resque.after_fork = Proc.new { ActiveRecord::Base.establish_connection }
OR
Resque.before_fork = Proc.new { ActiveRecord::Base.establish_connection }
However, I have done this (and tried any other thing I can think of) and I'm still getting the same error. I have run the code with only before_fork and only after_fork to no avail.
I am also using APN_sender to send apple push notifications. The workers for these have had no problems (but I'm calling default Heroku workers to do these rather than Resque workers).
Here are my relevant files, please help! This is my first SO question as well.. apologies if it's not done perfectly.
#config/resque.rb
after_fork do |server, worker, resque|
logger.info("Got to after_fork in resque.rb config file")
defined?(ActiveRecord::Base) and
ActiveRecord::Base.establish_connection
end
--------------------
#config/initializers/resque.rb
require 'resque'
require 'resque/server'
heroku_environments = ["staging", "production"]
rails_root = ENV['RAILS_ROOT'] || File.dirname(__FILE__) + '/../..'
rails_env = ENV['RAILS_ENV'] || 'development'
#confusing wording.. determines whether or not we are in Production, tested and working
unless heroku_environments.include?(rails_env)
resque_config = YAML.load_file(rails_root + '/config/resque.yml')
Resque.redis = resque_config[rails_env]
else
uri = URI.parse(ENV["REDISTOGO_URL"] || "redis://localhost:6379/")
Resque.redis = Redis.new(:host => uri.host, :port => uri.port, :password => uri.password)
end
----------------------
#resque.rake
require 'resque/tasks'
task "resque:setup" => :environment do
ENV['QUEUE'] = '*'
Resque.after_fork = Proc.new { ActiveRecord::Base.establish_connection }
Resque.before_fork = Proc.new { ActiveRecord::Base.establish_connection }
end
task "apn:setup" => :environment do
ENV['QUEUE'] = '*'
end
desc "Alias for resque:work (To run workers on Heroku)"
task "jobs:work" => "resque:work"
desc "Alias for apn:work (To run workers on Heroku)"
task "jobs:work" => "apn:work"
-----------------------
#Sending the email
#event.guests.each do |g|
Resque.enqueue(MailerCallback, "Notifier", :time_change, #event.id, g.id)
end
I am running one Heroku worker and one Resque worker from the Heroku ps:scale command. As I said, the first email sends error-free, and then any emails after get the above error.
Thanks in advance!
Mike
OK, so couldn't figure this out...
Switched over to delayed job and it's working much better. A little frustrating that I can't send push notifications with apn_sender anymore, but apn_on_rails is working fine.
Weird problem, still curious...