*****************************************************************
[composite:osapi_compute]
use = call:nova.api.openstack.urlmap:urlmap_factory
/: oscomputeversions
/v1.1: openstack_compute_api_v2
/v2: openstack_compute_api_v2
[composite:openstack_compute_api_v2]
use = call:nova.api.auth:pipeline_factory
noauth = faultwrap sizelimit noauth ratelimit osapi_compute_app_v2
keystone = faultwrap sizelimit authtoken keystonecontext ratelimit osapi_compute_app_v2
keystone_nolimit = faultwrap sizelimit authtoken keystonecontext osapi_compute_app_v2
[app:osapi_compute_app_v2]
paste.app_factory = nova.api.openstack.compute:APIRouter.factory
******************************************************************
1) #create wsgi service
server = service.WSGIService('osapi_compute') --->
2)# load WSGI applications from paste configurations
wsgi.Loader()
self.app = self.loader.load_app(name) --->
3) # urlmap_factory,
for path, app_name in local_conf.items():
path = paste.urlmap.parse_path_expression(path)
app = loader.get_app(app_name, global_conf=global_conf) --->
urlmap[path] = app
4) #pipeline_factory
app = loader.get_app(pipeline[-1]) -->
filters.reverse()
for filter in filters:
app = filter(app)
return app
这里的重要之处在于，通过app = filter(app) 将后面的filter app作为前面的middleware的构造初始化，从而形成整个call stack， 比如
keystone = faultwrap sizelimit authtoken keystonecontext ratelimit osapi_compute_app_v2
app = ratelimit(osapi_compute_app_v2)
这样ratelimit middleware的就会call osapi_compute_app_v2，
所以 3) 返回了第一个middleware的urlmap，从而 WSGI server可以在接收到/v2/***
进入middleware的 call stack ，最终根据APIRouter的url和controller的映射配置，进行API 的request的具体处理。
5) #APIRouter
class APIRouter(nova.api.openstack.APIRouter):
ExtensionManager = extensions.ExtensionManager
nova.api.openstack.APIRouter --->
if ext_mgr is None:
if self.ExtensionManager:
ext_mgr = self.ExtensionManager() --->
else:
raise Exception(_("Must specify an ExtensionManager class"))
mapper = ProjectMapper()
self.resources = {}
self._setup_routes(mapper, ext_mgr) (6)
self._setup_ext_routes(mapper, ext_mgr) (7)
self._setup_extensions(ext_mgr) (8)
:ExtensionManager
self._load_extensions() --->
load_extension() --->
:ExtensionDescriptor
ext_mgr.register(self)
:ExtensionManager
def _load_extensions(self):
"""Load extensions specified on the command line."""
extensions = list(self.cls_list)
for ext_factory in extensions:
try:
self.load_extension(ext_factory)
except Exception as exc:
LOG.warn(_('Failed to load extension %(ext_factory)s: '
'%(exc)s') % locals())
After all extensions have been loaded, (those extensions are configured in nova.conf, for example,
osapi_compute_extension=nova.api.openstack.compute.contrib.standard_extensions
6) self._setup_routes(mapper, ext_mgr)
for openstack core api, setup necessary routes
7) self._setup_ext_routes(mapper, ext_mgr)
for each resource in ext_mgr.get_resources()
wsgi_resource = wsgi.Resource(resource.controller,
inherits=inherits)
:Resource
self.register_actions(controller)
# Generate routes for a controller resource
# mapping a resource is about handling creating, viewing, and editing that resource
mapper.resource(resource.collection, resource.collection, **kargs)
8) self._setup_extensions(ext_mgr)
for each extension in returned ControllerExtension objects
get_controller_extensions
...
resource = self.resources[collection]
resource.register_actions(controller)
resource.register_extensions(controller)
:ExtensionDescriptor (append here for easy reference)
def get_resources(self):
"""List of extensions.ResourceExtension extension objects.
Resources define new nouns, and are accessible through URLs.
"""
resources = []
return resources
def get_controller_extensions(self):
"""List of extensions.ControllerExtension extension objects.
Controller extensions are used to extend existing controllers.
"""
controller_exts = []
return controller_exts
9) manage a WSGI server, serving a WSGI application --->
self.server = wsgi.Server(name,
self.app,
host=self.host,
port=self.port)
service.serve(server, workers=server.workers)
