Open Stack

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 10/10/14 20:21, Joshua Harlow wrote:
> On Oct 10, 2014, at 3:13 AM, Ihar Hrachyshka <ihrachys at redhat.com>
> wrote:
> 
> On 09/10/14 21:29, Mike Bayer wrote:
>>>> So so far, everyone seems really positive and psyched about
>>>> the proposal.
>>>> 
>>>> It looks like providing some options for how to use would be
>>>> best, that is provide decorators and context managers.
>>>> 
>>>> Now the thing with the context manager, it can be as I
>>>> stated:
>>>> 
>>>> with sql.reader() as session:
>>>> 
>>>> or we can even have that accept the “context”:
>>>> 
>>>> with sql.reader(context) as session:
>>>> 
>>>> The latter again avoids having to use thread locals.
>>>> 
>>>> But can I get a feel for how comfortable we are with using
>>>> thread local storage to implement this feature?   I had
>>>> anticipated people wouldn’t like it because it’s kind of a
>>>> “global” object, even though it will be hidden behind this
>>>> facade (of course CONF is global as is sys.modules, and I
>>>> think it is fine).     If I just use a tlocal, this whole
>>>> thing is pretty simple.
> 
> Won't the approach conflict with eventlet consumers that for some 
> reason do not patch thread module, or do not patch it early enough?
> I guess in that case we may end up with mixed contexts.
> 
>> Eck, this is not something people should be doing (not patching
>> the thread module).
> 
>> Example for why @
>> https://gist.github.com/harlowja/9c35e443dfa136a4f965 (run that
>> and see your program lock up).
> 
>> Once you accept/use eventlet, u shouldn't really be accepted half
>> of it, otherwise there be dragons there; especially since
>> openstack doesn't control what external library code does inside
>> those libraries (and rightfully so it shouldn't), and if any
>> library does anything like that gist code, the whole application
>> will lock up...

Real life sucks. In Neutron, we have at least two files that patch
only some modules (though all of them patch threading). [Not that I
mean that those files should not patch the whole library set; I have
no defined view on the matter.]

Though I agree that consumers with some modules unpatched are not
something that should be seriosly considered, we're still left with
consumers that do not patch their libraries early enough. We had an
issue with tlocal when porting Neutron to oslo.messaging that resulted
in the following patch: https://review.openstack.org/#/c/96791/ The
issue showed up in very weird way hard to debug, so if possible, I
would try to avoid similar situations with oslo.db.

/Ihar
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2.0.22 (Darwin)

iQEcBAEBCgAGBQJUODkZAAoJEC5aWaUY1u57QrcIANLtMi1jRrfHnwV2xnwYGPOl
QwZWDG6i/cdnzqrwGz+0chqRxu4KWZB5gAfMR4/ralHvGGbaa1vGp9wtzsfIJTB1
RuyE7XKXUX4rU3WoAOB7F3uF1WzFpI8ourunBG4OCE0t0YT89z9mCLfOTZNNKGxH
OE89n4pnyMd3GXGtoxVINyA6pqQotYuHKG6o6LbhmTZ1UoL3P3rO+Onb1Ma2BF6Z
VO84smdMnOC3yVtv4TGwrW5iBoU3RwOqxooOg4g5Jam6D++kzCKrPT9wdY7TNe+C
A6IvMlad0RbygkhrLyWIWXaqCAg5VzzDvk4z6dywVkmoWS7V861mjiV4unbr0X8=
=5ovh
-----END PGP SIGNATURE-----