Question

I have site on Pyramid framework and want to cache with memcached. For testing reasons I’ve used memory type caching and everything was OK. I’m using pyramid_beaker package.
Here is my previous code (working version).

In .ini file

cache.regions = day, hour, minute, second
cache.type = memory
cache.second.expire = 1
cache.minute.expire = 60
cache.hour.expire = 3600
cache.day.expire = 86400

In views.py:

from beaker.cache import cache_region

@cache_region('hour')
def get_popular_users():
    #some code to work with db
return some_dict

The only .ini settings I’ve found in docs were about working with memory and file types of caching. But I need to work with memcached.

First of all I’ve installed package memcached from Ubuntu official repository and also python-memcached to my virtualenv.

In .ini file I’ve replaced cache.type = memory -> cache.type = memcached. And I’ve got next error:

beaker.exceptions.MissingCacheParameter

MissingCacheParameter: url is required

What am I doing wrong?

Thanks in advance!

Answer 1

So, using the TurboGears documentation as a guide, what settings do you have for the url?

[app:main]
beaker.cache.type = ext:memcached
beaker.cache.url = 127.0.0.1:11211
# you can also store sessions in memcached, should you wish
# beaker.session.type = ext:memcached
# beaker.session.url = 127.0.0.1:11211

It looks to me as if memcached requires a url to initialize correctly:

def __init__(self, namespace, url=None, data_dir=None, lock_dir=None, **params):
    NamespaceManager.__init__(self, namespace)

    if not url:
        raise MissingCacheParameter("url is required") 

I am not really sure why the code allows url to be optional (defaulting to None) and then requires it. I think it would have been simpler just to require the url as an argument.

---

Later: in response to your next question:

when I used cache.url I’ve got next error: AttributeError:
‘MemcachedNamespaceManager’ object has no attribute ‘lock_dir’

I’d say that the way I read the code below, you have to provide either lock_dir or data_dir to initialize self.lock_dir:

    if lock_dir:
        self.lock_dir = lock_dir
    elif data_dir:
        self.lock_dir = data_dir + "/container_mcd_lock"
    if self.lock_dir:
        verify_directory(self.lock_dir)

You can replicate that exact error using this test code:

class Foo(object):
    def __init__(self, lock_dir=None, data_dir=None):
        if lock_dir:
            self.lock_dir = lock_dir
        elif data_dir:
            self.lock_dir = data_dir + "/container_mcd_lock"
        if self.lock_dir:
            verify_directory(self.lock_dir)

f = Foo()

It turns out like this:

>>> f = Foo()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 7, in __init__
AttributeError: 'Foo' object has no attribute 'lock_dir'