Interference with other network code about moto

[lucian1900]

Hmm, it doesn't help to create the redis connection outside the @mock_s3, since it (or httpretty?) monkey patches the socket module anyway.

[tponthieux]

I'm seeing something similar, using @mock_sqs seems to prevent me from being able to use httpretty at all.

[spulec]

Hmmm, this should all be fixed with gabrielfalcao/HTTPretty#34.

Currently moto will not work properly with httpretty since moto runs a modified version of httpretty. I've submitted pull requests to get it so that moto can use a non-custom version of httpretty. It's going to be difficult to do anything until those get pulled in.

[spulec]

I think this should be fixed now that we're back on the upstream version of HTTPretty. Can you confirm?

[lucian1900]

It appears to work with redis in my lock tests with moto.mock_s3. Sadly, it would appear other changes in moto have broken the rest of my tests so I can't be absolutely sure, but that's a separate issue :)

[spulec]

Ah, damn. Sorry to hear that.

I'm going to close this then. Do you mind opening some other issues when you get a minute?

[lucian1900]

Sure, I'll investigate that separately. I haven't given it much attention lately since the S3-using bits on my project are pretty much done.

[cgwaldman]

Hi
I'm seeing the "moto breaks redis" problem discussed above :(

Moto version:
from git, commit d8d55dc Sat Jan 24 08:59:14 2015 -0500

httpretty.version
u'0.8.0'

[spulec]

This is almost certainly an issue with HTTPretty. Can you narrow down a test case and open an issue there if that is true?

[cgwaldman]

I just opened gabrielfalcao/HTTPretty#212

also noted a related issue gabrielfalcao/HTTPretty#194 which has been untouched for 3 months

[cgwaldman]

Given that "HTTPretty monkey patches Python's socket core module", I don't think I'm going to be able to use moto. Pity.

[JackDanger]

After finding folks choosing to avoid the use of Moto and HTTPretty due to the socket monkey-patching (and finding myself unable to do so) I decided to bypass HTTPretty just inside my Redis/Memcached clients. Here's how I did it:

    class BypassHTTPretty():
        def __enter__(self):
            self.reenable_later = httpretty.is_enabled()
            if httpretty.is_enabled():
                httpretty.disable()

        def __exit__(self, type, value, traceback):
            if self.reenable_later:
                httpretty.enable()

    def redis_set(key, value):
        with BypassHTTPretty():
          redis.set(key, value)

Also posted here: gabrielfalcao/HTTPretty#212

[spulec]

Sorry about that. I'm working to move away from socket monkey-patching in the very near future.

Another option for the time being is to use Moto in standalone server mode: https://github.com/spulec/moto#stand-alone-server-mode