Re: [Openstack] Patched euca-tools to work w/ keystone
+1 BTW, the where are the backports committed? In the stable/diablo branch in github? Enviado desde mi iPad El 30/10/2011, a las 22:21, Leandro Reox leandro.r...@gmail.com escribió: Thanks a lot Ziad, will be cool to ge it in the back ports. Regards On Fri, Oct 28, 2011 at 5:11 PM, Ziad Sawalha ziad.sawa...@rackspace.com wrote: I think this was it: https://github.com/openstack/keystone/commit/2bb474331d73e7c6d2a507cb097c50 cfe65ad6b6 Will try to get it in the back ports. Z On 10/28/11 1:57 PM, Razique Mahroua razique.mahr...@gmail.com wrote: Hey, I never found out, in fact I only recall some mails exchange on a mailing list, basically, there are two lines to change into : /usr/local/lib/python2.6/dist-packages/keystone-1.0-py2.6.egg/keystone/mid dleware/ec2_token.py # o = urlparse(FLAGS.keystone_ec1_url) o = urlparse(FLAGS.keystone_ec2_url) and : # token_id = result['auth']['token']['id'] token_id = result['access']['token']['id'] Regards, Razique Le 28 oct. 2011 à 20:04, Leandro Reox a écrit : Hi guys, Anyone got the link to download the patched euca tools that work with keystone auth tokens to query the nova api ? Regards Lean ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] swift ring builder web service
Where can I find the spec for Swift blueprint, ring builder web service. In particular, I am looking for the definitions of new Swift RESTful API used by the web UI to perform listing, addition, deletion, modification, and rebalancing of the devices in the ring. And if the ring builder web service does not use RESTful API to communicate with Swift, what is the mechanism being used. Thanks Pauline___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] keystone Endpoint schema
What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
Well, If you need to specify a type when adding an endpointTemplate, then keystone should be smart enough to identify the type given and only accept the number of URLs needed for such type of service. Marcelo Martins Openstack-swift btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ On Oct 31, 2011, at 1:40 PM, Joseph Heck wrote: That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] Is there a reason Nova doesn't use scoped sessions in sqlalchemy ?
Hi Folks,
We've been looking into a problem which looks a lot like:
https://bugs.launchpad.net/nova/+bug/855660
2011-10-21 14:13:31,035 ERROR nova.api [5bd52130-d46f-4702-b06b-9ca5045473d7
smokeuser smokeproject] Unexpected error raised: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE: Traceback (most recent call last):
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/__init__.py, line 363, in
__call__
(nova.api): TRACE: result = api_request.invoke(context)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/apirequest.py, line 90, in
invoke
(nova.api): TRACE: result = method(context, **args)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1195, in
describe_instances
(nova.api): TRACE: instance_id=instance_id)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1204, in
_format_describe_instances
(nova.api): TRACE: return {'reservationSet': self._format_instances(context,
**kwargs)}
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1309, in
_format_instances
(nova.api): TRACE: if fixed['network'] and use_v6:
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/db/sqlalchemy/models.py, line 76, in
__getitem__
(nova.api): TRACE: return getattr(self, key)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 163, in
__get__
(nova.api): TRACE: instance_dict(instance))
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 383, in
get
(nova.api): TRACE: value = callable_(passive=passive)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/strategies.py, line 595, in
__call__
(nova.api): TRACE: (mapperutil.state_str(state), self.key)
(nova.api): TRACE: DetachedInstanceError: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE:
As far as we can see the problem seems to be related to some conflict between
multiple threads in the same API server instance and lazy loading of some part
of the object.
Looking at the sqlalchemy documentation it seems to strongly suggest that when
used from multi-threaded WSGI applications that scoped_sessions should be used
(I'm not clear on the details but it seems that this effectively makes lazy
load operations thread safe).However whilst this fixes the problem it has a
bad effect on the unit tests - in particular it seems to upset all of the DB
migration code used in the unit tests.
So does anyone know if there was an explicit decision / reason not to use
scoped_sessions in Nova ?
Thanks,
Phil
PS: The other possible fix we've found is to change sqlalchemy/models.py so
that the associations are explicitly set to use eager load - which also seems
to fix the problem but feels like a more clumsy way to go about it. Any
thoughts on that would also be appreciated ?
___
Mailing list: https://launchpad.net/~openstack
Post to : openstack@lists.launchpad.net
Unsubscribe : https://launchpad.net/~openstack
More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
It should require/accept the number of URLs that is required by the type of service one is adding. For example, swift only has public and localnet storage URLs. No admin URL. So, regardless if one is using keystone-manage or not (not sure what else one can use, Rest calls maybe ? ), it should only accept what the service type requires. case type: swift) try endpointTemplate add [region] [service] [public_url] [internal_url] [enabled] [is_global] except Failed with improper number of arguments show_some_help() nova) keystone) ... another-service) whatever_else) ... Marcelo Martins Openstack-swift btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ On Oct 31, 2011, at 1:52 PM, Joseph Heck wrote: Can you provide an example? I think you're asserting that you'd like the keystone-manage command to not require 3 different URLs when they don't exist separately, is that correct? -joe On Oct 31, 2011, at 11:45 AM, Marcelo Martins wrote: Well, If you need to specify a type when adding an endpointTemplate, then keystone should be smart enough to identify the type given and only accept the number of URLs needed for such type of service. Marcelo Martins Openstack-swift btorch...@zeroaccess.org On Oct 31, 2011, at 1:40 PM, Joseph Heck wrote: That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
Can you provide an example? I think you're asserting that you'd like the keystone-manage command to not require 3 different URLs when they don't exist separately, is that correct? -joe On Oct 31, 2011, at 11:45 AM, Marcelo Martins wrote: Well, If you need to specify a type when adding an endpointTemplate, then keystone should be smart enough to identify the type given and only accept the number of URLs needed for such type of service. Marcelo Martins Openstack-swift btorch...@zeroaccess.org On Oct 31, 2011, at 1:40 PM, Joseph Heck wrote: That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Is there a reason Nova doesn't use scoped sessions in sqlalchemy ?
On Mon, Oct 31, 2011 at 3:25 PM, Day, Phil philip@hp.com wrote:
Hi Folks,
We’ve been looking into a problem which looks a lot like:
https://bugs.launchpad.net/nova/+bug/855660
2011-10-21 14:13:31,035 ERROR nova.api [5bd52130-d46f-4702-b06b-9ca5045473d7
smokeuser smokeproject] Unexpected error raised: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE: Traceback (most recent call last):
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/__init__.py, line 363, in
__call__
(nova.api): TRACE: result = api_request.invoke(context)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/apirequest.py, line 90, in
invoke
(nova.api): TRACE: result = method(context, **args)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1195, in
describe_instances
(nova.api): TRACE: instance_id=instance_id)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1204, in
_format_describe_instances
(nova.api): TRACE: return {'reservationSet': self._format_instances(context,
**kwargs)}
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1309, in
_format_instances
(nova.api): TRACE: if fixed['network'] and use_v6:
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/db/sqlalchemy/models.py, line 76, in
__getitem__
(nova.api): TRACE: return getattr(self, key)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 163,
in __get__
(nova.api): TRACE: instance_dict(instance))
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 383,
in get
(nova.api): TRACE: value = callable_(passive=passive)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/strategies.py, line 595,
in __call__
(nova.api): TRACE: (mapperutil.state_str(state), self.key)
(nova.api): TRACE: DetachedInstanceError: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE:
Have you tried backporting the fix for the bug you mentioned above to
your local branch? Did that have any affect?
Thanks!
-jay
___
Mailing list: https://launchpad.net/~openstack
Post to : openstack@lists.launchpad.net
Unsubscribe : https://launchpad.net/~openstack
More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Is there a reason Nova doesn't use scoped sessions in sqlalchemy ?
All of the workers are single-threaded, so I'm not sure that scoped sessions
are really necessary.
We did however decide that objects from the db layer are supposed to be simple
dictionaries. We currently allow nested dictionaries to optimize joined
objects. Unfortunately we never switched to sanitizing data from sqlalchemy,
and instead we make the sqlalchemy objects provide a dictionary-like interface
and pass the object itself.
The issue that you're seeing is because network wasn't properly 'joinedload'ed
in the initial query, and because the data is not sanitized, sqlalchemy tries
to joinedload, but the session has been terminated. If we had sanitized data,
we would get a more useful error like a key error when network is accessed. The
current solution is to add the proper joinedload.
One of the goals of the nova-database team is to do the necessary data
sanitization and to remove as many of the joinedloads as possible (hopefully
all of them).
Vish
On Oct 31, 2011, at 12:25 PM, Day, Phil wrote:
Hi Folks,
We’ve been looking into a problem which looks a lot like:
https://bugs.launchpad.net/nova/+bug/855660
2011-10-21 14:13:31,035 ERROR nova.api [5bd52130-d46f-4702-b06b-9ca5045473d7
smokeuser smokeproject] Unexpected error raised: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE: Traceback (most recent call last):
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/__init__.py, line 363, in
__call__
(nova.api): TRACE: result = api_request.invoke(context)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/apirequest.py, line 90, in
invoke
(nova.api): TRACE: result = method(context, **args)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1195, in
describe_instances
(nova.api): TRACE: instance_id=instance_id)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1204, in
_format_describe_instances
(nova.api): TRACE: return {'reservationSet': self._format_instances(context,
**kwargs)}
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/api/ec2/cloud.py, line 1309, in
_format_instances
(nova.api): TRACE: if fixed['network'] and use_v6:
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/nova/db/sqlalchemy/models.py, line 76, in
__getitem__
(nova.api): TRACE: return getattr(self, key)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 163, in
__get__
(nova.api): TRACE: instance_dict(instance))
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/attributes.py, line 383, in
get
(nova.api): TRACE: value = callable_(passive=passive)
(nova.api): TRACE: File
/usr/lib/python2.7/dist-packages/sqlalchemy/orm/strategies.py, line 595, in
__call__
(nova.api): TRACE: (mapperutil.state_str(state), self.key)
(nova.api): TRACE: DetachedInstanceError: Parent instance FixedIp at
0x4e74490 is not bound to a Session; lazy load operation of attribute
'network' cannot proceed
(nova.api): TRACE:
As far as we can see the problem seems to be related to some conflict between
multiple threads in the same API server instance and lazy loading of some
part of the object.
Looking at the sqlalchemy documentation it seems to strongly suggest that
when used from multi-threaded WSGI applications that scoped_sessions should
be used (I’m not clear on the details but it seems that this effectively
makes lazy load operations thread safe).However whilst this fixes the
problem it has a bad effect on the unit tests – in particular it seems to
upset all of the DB migration code used in the unit tests.
So does anyone know if there was an explicit decision / reason not to use
scoped_sessions in Nova ?
Thanks,
Phil
PS: The other possible fix we’ve found is to change sqlalchemy/models.py so
that the associations are explicitly set to use eager load – which also seems
to fix the problem but feels like a more clumsy way to go about it. Any
thoughts on that would also be appreciated ?
___
Mailing list: https://launchpad.net/~openstack
Post to : openstack@lists.launchpad.net
Unsubscribe : https://launchpad.net/~openstack
More help : https://help.launchpad.net/ListHelp
___
Mailing list: https://launchpad.net/~openstack
Post to : openstack@lists.launchpad.net
Unsubscribe : https://launchpad.net/~openstack
More help : https://help.launchpad.net/ListHelp
[Openstack] Gflags / conf - common?
Hi all, I was wondering if there is any plans in essex to standardize either using gflags or using configuration files for these types of settings. One of the complaints that I receive a lot with gflags is that by including a python file, u automatically inject all of its flags (even if they are not used) into gflags (since its global). Thus say u are just using the nova-compute run time, but that itself includes say flags.py which itself seems to be a common area for flags that may or may not be used by that runtime. Similarly if a file is imported has say 1 method used by the calling code but itself defines 10 flags (for its components) then those 10 flags get injected. This makes it very confusing to figure out what should be set (or what could be set). Has there been any thought on fixing this (or making a standard recommendation that subprojects can follow) that would avoid this problem? I could imagine fixes being in the code structure itself (having said 1 method stated above not be in a file what pulls in other code that defines 10 flags) or another type of configuration mechanism? I think this was mentioned at the conference, but not sure what came out of that :-) -Josh ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Gflags / conf - common?
Hi! GFlags has now been removed, AFAIK. The flags module has an optparse-based emulator for GFlags to ease transition for Nova joining the rest of the OpenStack core project implementations' use of standard config files/Paste.Deploy. Cheers, -jay On Mon, Oct 31, 2011 at 5:08 PM, Joshua Harlow harlo...@yahoo-inc.com wrote: Hi all, I was wondering if there is any plans in essex to standardize either using gflags or using configuration files for these types of settings. One of the complaints that I receive a lot with gflags is that by including a python file, u automatically inject all of its flags (even if they are not used) into gflags (since its global). Thus say u are just using the nova-compute run time, but that itself includes say “flags.py” which itself seems to be a common area for flags that may or may not be used by that runtime. Similarly if a file is imported has say 1 method used by the calling code but itself defines 10 flags (for its components) then those 10 flags get injected. This makes it very confusing to figure out what should be set (or what could be set). Has there been any thought on fixing this (or making a standard recommendation that subprojects can follow) that would avoid this problem? I could imagine fixes being in the code structure itself (having said 1 method stated above not be in a file what pulls in other code that defines 10 flags) or another type of configuration mechanism? I think this was mentioned at the conference, but not sure what came out of that :-) -Josh ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
The list of URLs comes from what we have historically done at Rackspace and the conversations had in OpenStack about a management/admin API. I agree that not all services need those three. And some may want to create additional ones. You mention type below. Not to be confused with the serviceType (like compute, identity, image-service, object-store, etc...). Are you proposing an EndpointType (maybe admin, public, private, etc..)? That does seem like a more flexible approach. It would help to have some well-known types, such as: - public: Internet-accessible - admin: private, with elevated-privilege calls available - internal: provides a high bandwidth, low latency, unmetered endpoint Thoughts? Z On Oct 31, 2011, at 2:17 PM, Marcelo Martins btorch...@zeroaccess.orgmailto:btorch...@zeroaccess.org wrote: It should require/accept the number of URLs that is required by the type of service one is adding. For example, swift only has public and localnet storage URLs. No admin URL. So, regardless if one is using keystone-manage or not (not sure what else one can use, Rest calls maybe ? ), it should only accept what the service type requires. case type: swift) try endpointTemplate add [region] [service] [public_url] [internal_url] [enabled] [is_global] except Failed with improper number of arguments show_some_help() nova) keystone) ... another-service) whatever_else) ... Marcelo Martins Openstack-swift btorch...@zeroaccess.orgmailto:btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ On Oct 31, 2011, at 1:52 PM, Joseph Heck wrote: Can you provide an example? I think you're asserting that you'd like the keystone-manage command to not require 3 different URLs when they don't exist separately, is that correct? -joe On Oct 31, 2011, at 11:45 AM, Marcelo Martins wrote: Well, If you need to specify a type when adding an endpointTemplate, then keystone should be smart enough to identify the type given and only accept the number of URLs needed for such type of service. Marcelo Martins Openstack-swift btorch...@zeroaccess.orgmailto:btorch...@zeroaccess.org On Oct 31, 2011, at 1:40 PM, Joseph Heck wrote: That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.orgmailto:btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.netmailto:openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] OpenStack and OpenFlow
On Fri, Oct 28, 2011 at 10:58 PM, Yoshisato Ushio yoshisato.us...@gmail.com wrote: Hello Folks, I am interested in the use of OpenFlow technology with OpenStack. But honestly speaking, I have no deep expertise in OpenStack.Then let me ask questions. Q1) In order to use openflow features with OpenStack, such as, Open vSwitch or OpenFlow capable HW switches, can I assume that all I should do for this is to develop a quantum plugin module for openflow under nova manager ? Hi Yoshisato, If you are trying to enable the use of OpenStack with a new network switching technology, then yes, developing a Quantum plugin is the right way to go. It may be possible to leverage work done by others writing similar plugins though. Since our developer documentation as-of-yet non-existent, I whipped up a quick wiki page with some advice on developing a plugin: http://wiki.openstack.org/QuantumDevelopment . Others should feel free to expand on this page. Also, just to clarify one thing though: Quantum is actually a separate codebase + runtime process from Nova, so you shouldn't need to modify Nova itself. Q2) Are there any open-source codes of quantum plugin for openflow ? I doubt that there will be one canonical openflow plugin for Quantum, as OpenFlow itself is really just a protocol to talk to a (virtual or physical) switch's forwarding table. There are many different ways to build a Quantum plugin that uses OpenFlow, each of which make different trade-offs in terms of scalability, fault-tolerance, complexity, feature-richness, etc. You can probably start out by taking a look at some of the existing plugins and seeing what ideas you can borrow with respect to your own implementation. Feel free to send email to the netstack list if you have more detailed questions. Dan Any advice would be greatly appreciated, Yoshisato Ushio ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp -- ~~~ Dan Wendlandt Sr. Product Manager Nicira Networks: www.nicira.com cell: 650-906-2650 twitter: danwendlant ~~~ ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] [GLANCE] OpenStack Images API 2.0 Proposal Updated
Hi all, Just a thank you to all the folks that have provided feedback on the proposed OpenStack Images API 2.0 proposal: https://docs.google.com/document/d/16QmRFzKyyphhI20muxpx7sNhUUHpzbQ7XEUMVTYbmRw/edit I've addressed the feedback up until now and invite previous reviewers and everyone else to inspect the updated document. Anyone may provide feedback on the document simply by right-clicking on an area and clicking Comment. Also, I'm working on creating all the blueprints associated with getting an implementation of the proposed OpenStack Images API 2.0 into Glance in the Essex release cycle: https://blueprints.launchpad.net/glance/+spec/api-2 Feel free to provide any feedback you might have on those blueprints by replying to this email, and if you're interested in contributing to the implementation efforts, please do shoot me an email :) Cheers! -jay ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Gflags / conf - common?
Great. Thx! On 10/31/11 2:42 PM, Jay Pipes jaypi...@gmail.com wrote: Hi! GFlags has now been removed, AFAIK. The flags module has an optparse-based emulator for GFlags to ease transition for Nova joining the rest of the OpenStack core project implementations' use of standard config files/Paste.Deploy. Cheers, -jay On Mon, Oct 31, 2011 at 5:08 PM, Joshua Harlow harlo...@yahoo-inc.com wrote: Hi all, I was wondering if there is any plans in essex to standardize either using gflags or using configuration files for these types of settings. One of the complaints that I receive a lot with gflags is that by including a python file, u automatically inject all of its flags (even if they are not used) into gflags (since its global). Thus say u are just using the nova-compute run time, but that itself includes say flags.py which itself seems to be a common area for flags that may or may not be used by that runtime. Similarly if a file is imported has say 1 method used by the calling code but itself defines 10 flags (for its components) then those 10 flags get injected. This makes it very confusing to figure out what should be set (or what could be set). Has there been any thought on fixing this (or making a standard recommendation that subprojects can follow) that would avoid this problem? I could imagine fixes being in the code structure itself (having said 1 method stated above not be in a file what pulls in other code that defines 10 flags) or another type of configuration mechanism? I think this was mentioned at the conference, but not sure what came out of that :-) -Josh ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] [Nova] [Keystone] Nova projects (Diablo)
Hi, 2011/11/01 1:07 Nguyen, Liem Manh liem_m_ngu...@hp.com: Cool, will this patch be available off stable/Diablo eventually, or would it be in Essex only? I hope it will be a part of Diablo stable updates and Essex. Regards, Akira Yoshiyama NEC Corp. ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] keystone Endpoint schema
Hi Ziad, Sorry, that was my mistake. I meant to have case service.name: on that pseudocode and not type. I wasn't proposing any EndpointType and don't see how that would help. The way that I was thinking was, you can either have the services table pre-populated during keystone install/setup or have the user do it. Also provide information on the docs about the services that keystone currently support. The documentation would provide information to the user on how to add an endpointTemplate to a particular service. Perhaps this is a bit more clear: case service.name: openstack-swift) try endpointTemplate add [region] [service] [public_url] [internal_url] [enabled] [is_global] except Failed with improper number of argumentsswift) show_some_help() try openstack-compute) endpointTemplate add [region] [service] [public_url] [admin_url] [internal_url] [enabled] [is_global] openstack-swift) If one needs to add a generic/custom service (one that keystone does not know yet), perhaps keystone could be flexible enough to accept N numbers of URLs for this generic/custom service. But I think this is something more for the future. Marcelo Martins Openstack-swift btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ On Oct 31, 2011, at 4:42 PM, Ziad Sawalha wrote: The list of URLs comes from what we have historically done at Rackspace and the conversations had in OpenStack about a management/admin API. I agree that not all services need those three. And some may want to create additional ones. You mention type below. Not to be confused with the serviceType (like compute, identity, image-service, object-store, etc...). Are you proposing an EndpointType (maybe admin, public, private, etc..)? That does seem like a more flexible approach. It would help to have some well-known types, such as: - public: Internet-accessible - admin: private, with elevated-privilege calls available - internal: provides a high bandwidth, low latency, unmetered endpoint Thoughts? Z On Oct 31, 2011, at 2:17 PM, Marcelo Martins btorch...@zeroaccess.org wrote: It should require/accept the number of URLs that is required by the type of service one is adding. For example, swift only has public and localnet storage URLs. No admin URL. So, regardless if one is using keystone-manage or not (not sure what else one can use, Rest calls maybe ? ), it should only accept what the service type requires. case type: swift) try endpointTemplate add [region] [service] [public_url] [internal_url] [enabled] [is_global] except Failed with improper number of arguments show_some_help() nova) keystone) ... another-service) whatever_else) ... Marcelo Martins Openstack-swift btorch...@zeroaccess.org “Knowledge is the wings on which our aspirations take flight and soar. When it comes to surfing and life if you know what to do you can do it. If you desire anything become educated about it and succeed. “ On Oct 31, 2011, at 1:52 PM, Joseph Heck wrote: Can you provide an example? I think you're asserting that you'd like the keystone-manage command to not require 3 different URLs when they don't exist separately, is that correct? -joe On Oct 31, 2011, at 11:45 AM, Marcelo Martins wrote: Well, If you need to specify a type when adding an endpointTemplate, then keystone should be smart enough to identify the type given and only accept the number of URLs needed for such type of service. Marcelo Martins Openstack-swift btorch...@zeroaccess.org On Oct 31, 2011, at 1:40 PM, Joseph Heck wrote: That's just what it sees today - the only one of the service endpoints that uses all three (right now anyway) is Keystone itself. Can you share a different pattern that you're interested in seeing supported? -joe On Oct 31, 2011, at 9:46 AM, Marcelo Martins wrote: What makes keystone assume that all types of services will have [public_url] [admin_url] [internal_url] ? Marcelo Martins Openstack-swift btorch...@zeroaccess.org ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
