Two requests to one file are not done in parallel

[mmf1987]

BPO	1227480
Nosy	@birkenfeld, @josiahcarlson
Files	test1.py: test1.py - the little CGI script written in Python

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2005-06-30.08:49:54.000>
created_at = <Date 2005-06-25.19:24:48.000>
labels = ['interpreter-core']
title = 'Two requests to one file are not done in parallel'
updated_at = <Date 2005-06-30.08:49:54.000>
user = 'https://bugs.python.org/mmf1987'

bugs.python.org fields:

activity = <Date 2005-06-30.08:49:54.000>
actor = 'jafo'
assignee = 'none'
closed = True
closed_date = None
closer = None
components = ['Interpreter Core']
creation = <Date 2005-06-25.19:24:48.000>
creator = 'mmf1987'
dependencies = []
files = ['1728']
hgrepos = []
issue_num = 1227480
keywords = []
message_count = 11.0
messages = ['25632', '25633', '25634', '25635', '25636', '25637', '25638', '25639', '25640', '25641', '25642']
nosy_count = 4.0
nosy_names = ['georg.brandl', 'jafo', 'josiahcarlson', 'mmf1987']
pr_nums = []
priority = 'normal'
resolution = 'works for me'
stage = None
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue1227480'
versions = []

[mmf1987]

Hi!

I am using Python for CGI scripting. I had a little cgi
script (see the attached file).

I saved it as test1.py - and tried to load it over HTTP. So
everything is fine.

But now I saw a very strange behaviour: I sent two
parallel requests to
the script (for example by sending two requests in two
browser windows
almost at the same time).

My problem: The requests to the script seem NOT to be
done in parallel:
The first script prints "starting..." and after a short time
"finished". And then in the other browser window the
second request
shows "starting...".

But I expected that in both browser windows "starting"
will be shown
almost at the same time because the two requests are
done in parallel.

The strange behaviour goes on:
If copied the test1.py and saved it as test2.py. Now I
retried my test
and sent two requests - but to these two identical files
(that only
differ in their filename). The the two requests seem to be
done in
parallel...

(It tried my test on Fedora and Gentoo Linux, with
Python 2.2 and
Python 2.4, with different webservers (Apache and
others), and with
several different browsers - I always was the same.)

I think the behaviour will be the same if I use my script
on the command line...

Best regards,
Markus Franz

[jafo]

Logged In: YES
user_id=81797

There's no uploaded file! You have to check the
checkbox labeled "Check to Upload & Attach File"
when you upload a file.

Please try again.

(This is a SourceForge annoyance that we can do
nothing about. :-( )

[jafo]

Logged In: YES
user_id=81797

There's no uploaded file! You have to check the
checkbox labeled "Check to Upload & Attach File"
when you upload a file.

Please try again.

(This is a SourceForge annoyance that we can do
nothing about. :-( )

[jafo]

Logged In: YES
user_id=81797

There's no uploaded file! You have to check the
checkbox labeled "Check to Upload & Attach File"
when you upload a file.

Please try again.

(This is a SourceForge annoyance that we can do
nothing about. :-( )

[jafo]

Logged In: YES
user_id=81797

There's no uploaded file! You have to check the
checkbox labeled "Check to Upload & Attach File"
when you upload a file.

Please try again.

(This is a SourceForge annoyance that we can do
nothing about. :-( )

[jafo]

Logged In: YES
user_id=81797

There's no uploaded file! You have to check the
checkbox labeled "Check to Upload & Attach File"
when you upload a file.

Please try again.

(This is a SourceForge annoyance that we can do
nothing about. :-( )

[josiahcarlson]

Logged In: YES
user_id=341410

I doubt that this is a Python issue. It seems more likely
that your web servers are only willing to spawn a single CGI
process at any one time.

[jafo]

Logged In: YES
user_id=81797

I would tend to agree that it's unlikely to be a Python
issue, but I'd be willing to try to reproduce it given
further information.

[josiahcarlson]

Logged In: YES
user_id=341410

It's not a Python issue. I can run multiple copies of
Python in the same console. It's the
non-forking/non-threaded web server the OP is using.

bash-2.05b$ /cygdrive/c/python24/python -c \

"import time;time.sleep(1);import
sys;sys.stdout.write('hello\r\n');\
sys.stdout.flush();[i for i in xrange(10000000)];\
ss.stdout.write('goodbye\r\n');sys.stdout.flush()" &
[1] 2380
bash-2.05b$ /cygdrive/c/python24/python -c "import
time;time.sleep(1);import sy
s;sys.stdout.write('hello\r\n');\
sys.stdout.flush();[i for i in xrange(10000000)];\
sys.stdout.write('goodbye\r\n');sys.stdout.flush()" &
[2] 1468
bash-2.05b$ hello
hello
goodbye
goodbye

[1]- Done /cygdrive/c/python24/python -c
"import time;time.s
leep(1);import
sys;sys.stdout.write('hello\r\n');sys.stdout.flush();[i for i in
xrange(10000000)];sys.stdout.write('goodbye\r\n');sys.stdout.flush()"
[2]+ Done /cygdrive/c/python24/python -c
"import time;time.s
leep(1);import
sys;sys.stdout.write('hello\r\n');sys.stdout.flush();[i for i in
xrange(10000000)];sys.stdout.write('goodbye\r\n');sys.stdout.flush()"
bash-2.05b$

[birkenfeld]

Logged In: YES
user_id=1188172

What's the matter with the filenames, then? If the OP is
right and when the filenames differ, parallel is no problem,
is it a webserver problem too?

[jafo]

Logged In: YES
user_id=81797

This is definitely not a Python problem. If I run your
script from two terminal windows from Python directly, they
both finish at the same time. So, this is either something
in how your browser or proxy is sending requests (perhaps
there is some serialization going on in there), or something
with how your web server is handling the requests. I can't
say for sure, but it's unlikely to be an issue with Python.
I'm closing this ticket because of that.