Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Speaker Deck
PRO
Sign in
Sign up
for free
The Long Road To Asynchrony
Andrew Godwin
February 22, 2020
Programming
0
270
The Long Road To Asynchrony
My keynote from PyCon Belarus 2020.
Andrew Godwin
February 22, 2020
Tweet
Share
More Decks by Andrew Godwin
See All by Andrew Godwin
andrewgodwin
0
80
andrewgodwin
0
61
andrewgodwin
0
240
andrewgodwin
0
44
andrewgodwin
2
640
andrewgodwin
1
120
andrewgodwin
0
36
andrewgodwin
0
96
andrewgodwin
0
300
Other Decks in Programming
See All in Programming
line_developers_tw
0
500
anchorcable
1
120
yshrsmz
1
450
zsmb
2
110
akatsukinewgrad
0
180
ippey
0
170
kyoheig3
0
420
nbkouhou
0
890
o0h
PRO
1
1.5k
manfredsteyer
PRO
0
270
asumikan
0
280
akatsukinewgrad
0
190
Featured
See All Featured
roundedbygravity
241
21k
addyosmani
310
21k
rocio
155
11k
dougneiner
119
7.8k
maggiecrowley
8
400
pauljervisheath
196
15k
garrettdimon
287
110k
bryan
100
11k
ddemaree
274
31k
chriscoyier
145
19k
edds
56
9.3k
orderedlist
PRO
329
35k
Transcript
LONG ROAD ANDREW GODWIN // @andrewgodwin ASYNCHRONY TO THE
Andrew Godwin / @andrewgodwin Hi, I’m Andrew Godwin • Django
core developer • Worked on Migrations, Channels & Async • Once a Londoner, now from Denver, USA
Andrew Godwin / @andrewgodwin
Andrew Godwin / @andrewgodwin "Asynchronous Programming" What is it, really?
Andrew Godwin / @andrewgodwin Concurrent Programming The more general term
Andrew Godwin / @andrewgodwin Input Process Output Sequential execution
Andrew Godwin / @andrewgodwin Input Process Output Process Concurrent execution
Archive
Andrew Godwin / @andrewgodwin Shared use of a single resource
In this case, CPUs
Andrew Godwin / @andrewgodwin "Communicating Sequential Processes", C. A. R
Hoare
Andrew Godwin / @andrewgodwin func main() { messages := make(chan
string) go func() { messages <- "ping" }() msg := <-messages fmt.Println(msg) }
Andrew Godwin / @andrewgodwin Multiple processes multiprocessing Threads threading Event
loops asyncio / twisted
Andrew Godwin / @andrewgodwin Multiple processes scale best It's also
difficult and costs the most!
Andrew Godwin / @andrewgodwin Threads are unpredictable Also, the GIL
is our ever-present friend
Andrew Godwin / @andrewgodwin Event loops are a good compromise
They do require shared memory, though.
Andrew Godwin / @andrewgodwin Asynchronous ≈ Event loops Most of
the time!
Andrew Godwin / @andrewgodwin # Ready when a timer finishes
await asyncio.sleep(1) # Ready when network packets return await client.get("http://example.com") # Ready when the coroutine exits await my_function("hello", 64.2)
Andrew Godwin / @andrewgodwin Network/timer updates An event loop's flow
Select a ready task Run task Add new tasks to queue await
Andrew Godwin / @andrewgodwin Coroutines Time →
Andrew Godwin / @andrewgodwin How did we get here?
Andrew Godwin / @andrewgodwin 1998 threading module, Stackless Python 2002
Twisted 2006 Greenlets (later gevent, eventlet) 2008 multiprocessing module 2012 Tulip, PEP 3156 2014 asyncio module 2005 Coroutine-friendly generators (PEP 342)
Andrew Godwin / @andrewgodwin 2017 Django Channels 1.0 2018 Django
Channels 2.0 2019 DEP 9 (Async support) 2020 Async views land in Django
Andrew Godwin / @andrewgodwin No solution is perfect Everyone chooses
different tradeoffs
Andrew Godwin / @andrewgodwin Asyncio is based on yield from
Because it was prototyped in Python 2
Andrew Godwin / @andrewgodwin Can't tell if a function returns
a coroutine! There are standard hints, but no actual guaranteed way
Andrew Godwin / @andrewgodwin async def calculate(x): result = await
coroutine(x) return result # These both return a coroutine def calculate(x): result = coroutine(x) return result
Andrew Godwin / @andrewgodwin Can't have one function service both
How we got here makes sense, but it's still annoying sometimes.
Andrew Godwin / @andrewgodwin # Calls get.__call__ instance = MyModel.objects.get(id=3)
# Calls get.__call__ # and then awaits its result instance = await MyModel.objects.get(id=3)
Andrew Godwin / @andrewgodwin You have to namespace async functions
I really, really wish we didn't have to
Andrew Godwin / @andrewgodwin instance = MyModel.objects.get(id=3) instance = await
MyModel.objects.async.get(id=3)
Andrew Godwin / @andrewgodwin Completely different libraries! Even sleep() is
different.
Andrew Godwin / @andrewgodwin time.sleep ➞ asyncio.sleep requests ➞ httpx
psycopg2 ➞ aiopg WSGI ➞ ASGI Django ➞ Django?
Andrew Godwin / @andrewgodwin Django & Async
Andrew Godwin / @andrewgodwin Asyncio only benefits IO-bound code Code
that thrashes the CPU doesn't benefit at all
Andrew Godwin / @andrewgodwin We're adding async to some parts
The bits where it makes sense!
Andrew Godwin / @andrewgodwin But, you can't mix sync and
async So we have to have two parallel request paths
Andrew Godwin / @andrewgodwin WSGIHandler __call__ WSGI Server WSGIRequest BaseHandler
get_response URLs Middleware View __call__ HTTP protocol Socket handling Transfer encodings Headers-to-META Upload file wrapping GET/POST parsing Exception catching Atomic view wrapper Django 3.0 Request Flow
Andrew Godwin / @andrewgodwin WSGIHandler __call__ WSGI Server WSGIRequest BaseHandler
get_response URLs Middleware Async View __call__ ASGIHandler __call__ ASGI Server ASGIRequest Sync View __call__ Asynchronous request path Proposed async request flow
Andrew Godwin / @andrewgodwin WSGIHandler __call__ WSGI Server WSGIRequest URLs
Middleware View __call__ ASGIHandler __call__ ASGI Server ASGIRequest Asynchronous request path BaseHandler get_response_async BaseHandler get_response URLs Middleware Async View __call__ Implemented async request flow
Andrew Godwin / @andrewgodwin We have to work with what
we have I'd rather let people ship code than argue about perfection.
Andrew Godwin / @andrewgodwin Django's main job is safety It
matters more than anything else
Andrew Godwin / @andrewgodwin Deadlocks Livelocks Starvation Race conditions
Andrew Godwin / @andrewgodwin Coroutines & the GIL actually help!
You're saved from all the awful memory corruption bugs
Andrew Godwin / @andrewgodwin async def transfer_money(p1, p2): await get_lock()
# Code between awaits is atomic! subtract_money(p1) add_money(p2) await release_lock()
Andrew Godwin / @andrewgodwin You can still screw up a
lot Trust me, I have lived it while developing async
Andrew Godwin / @andrewgodwin async def log_message(m): await client.post("log-server", m)
result = calculate_result() log_message(m)
Andrew Godwin / @andrewgodwin async def log_message(m): await client.post("log-server", m)
result = calculate_result() await log_message(m)
Andrew Godwin / @andrewgodwin Async usability has a long way
to go But it is undoubtedly the future!
Andrew Godwin / @andrewgodwin Python is the language of pragmatism
If anyone can get it right, we can
Andrew Godwin / @andrewgodwin What does the future hold? Hopefully,
no GIL!
Andrew Godwin / @andrewgodwin Let's make async understandable Almost every
project could benefit, if we made it worth their time.
Thanks. Andrew Godwin @andrewgodwin // aeracode.org