Re: [Openstack] Providing packages for stable releases of OpenStack
On Tue, 2011-12-06 at 14:12 -0800, Duncan McGreggor wrote: On 06 Dec 2011 - 13:52, Duncan McGreggor wrote: On 06 Dec 2011 - 21:14, Thierry Carrez wrote: Tim Bell wrote: I'm not clear on who will be maintaining the stable/diablo branch. The people such as EPEL for RedHat systems need to have something with the appropriate bug fixes back ported. There are an increasing number of sites looking to deploy in production and cannot follow the latest development version. Agreed on the need, we discussed this at length during the design summit. The stable branches have been established and are maintained by the OpenStack Stable Branch Maintainers team. Currently this team is mostly made of distribution members (Ubuntu and Fedora/RedHat, mostly) collaborating on a single branch to avoid duplication of effort. See: https://launchpad.net/~openstack-stable-maint http://wiki.openstack.org/StableBranch Okay, I think this mostly addresses item #4 that I wanted to add to your summary, Thierry. I do have the following minor concerns, though: * that wiki page's summary (intro sentence) only specifically mentions Diablo; I'd like to see something along the lines of currently focusing on Diablo. If these processes evolve into a successful model, they will be applied to all future releases. Added. * the discussion on the page treats this as an experiment (this is good!), but I'd like to see a phrase alone the lines of if this experiment is successful, we will do X to ensure these processes become an official part of the workflow. Cleaned up the this is an experiment text a bit, it's gone beyond an experiment now I think. These are tiny things, but I think they will better set expectations and give more warm fuzzies to organizations thinking about deploying OpenStack in production environments, seeing that we're considering the long-term (given success of the experiment). In addition, I would like to emphasize Tim's point from earlier, though: it's not just packaging... Note that the stable-maint team has no involvement with upstream packaging, if that's what you're talking about. Cheers, Mark. ___ 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] Providing packages for stable releases of OpenStack
On Tue, 2011-12-06 at 10:11 -0800, Duncan McGreggor wrote: On 06 Dec 2011 - 14:28, Thierry Carrez wrote: So the general consensus so far on this discussion seems to be: (0) The 2011.3 release PPA bears false expectations and should be removed now. In the future, we should not provide such PPAs: 0-day packages for the release should be available from the last milestone PPA anyway. (1) OpenStack, as an upstream project, should focus on development rather than on providing a production-ready distribution. (2) We could provide daily builds from the stable/diablo branch for a variety of releases (much like what we do for the master branch), but those should be clearly marked not for production use and be best-effort only (like our master branch builds). (3) This should not prevent a group in the community from working on a project providing an openstack on Lucid production-ready distribution if they so wishes. This project would just be another distribution of OpenStack. This doesn't seem like enough to me. OpenStack isn't just a library; it's a fairly substantial collection of software and services, intended to be used as a product. If it can't be used as a product, what's the use? Someone made the suggestion that a new OpenStack group be started, one whose focus is on producing a production-ready, distribution-ready, release of the software. So can we add one more (need some help with wording, here...): This paragraph makes it sound like you think OpenStack upstream should produce production-ready binary packages? This is what we've agreed not to do, at least until some group of volunteers show that they can sustain the work involved. (4) OpenStack will accept and foster a new project, one that is not focused on development, but rather the distribution and it's general stability. This distro project will be responsible for advocating on behalf of various operating systems/distros/sponsoring vendors for bugs that affect performance and stability of OpenStack, or prevent an operating system from running OpenStack. This sounds like you want want OpenStack to focus more on the quality of its releases. This is generally accepted and you can join the QA team to help out: https://launchpad.net/~openstack-qa-team It also sounds like you want more co-ordination upstream between downstream distributions of OpenStack. I think there's already pretty good co-ordination. Is there any specific problem you've seen on that front? Cheers, Mark. ___ 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 Subteam Changes
On Tue, 2011-12-06 at 19:54 -0800, Vishvananda Ishaya wrote: Hello Everyone, The Nova subteams have now been active for a month and a half. Some things are going very well, and others could use a little improvement. To keep things moving forward, I'd like to make the following changes: 1) Weekly meeting for team leads. [..] 2) Closing down the team mailinglists. [..] 3) Closing teams. Some of the teams haven't really started having much activity. [..] Sounds good to me. In future, perhaps new teams should only be formed when there are two or more people actively working on stuff and need to co-ordinate? It was a little troubling at the summit that some discussions ended with agreement to start a team, but no obvious sense that anyone was going to step up and actively do the work. Cheers, Mark. ___ 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 Subteam Changes
Vishvananda Ishaya wrote: 2) *Closing down the team mailinglists.* Some of the lists have been a bit active, but I think the approach that soren has been using of sending messages to the regular list with a team [header] is a better approach. Examples: [db] Should we use zookeeper? [scaling] Plans for bursting I realize that this will lead to a little more noise on some of the channels, but I think it makes sure that we don't isolate knowledge too much Could we have a standardized list of headers ? Those would double as tags for Nova bugs, so the scaling team could see a list of bugs in their area by searching for the scaling tag. 3) *Closing teams.* Some of the teams haven't really started having much activity. I'm going to move these teams into a WANTED section on the teams page here: http://wiki.openstack.org/Teams For now, people can just discuss stuff on the main mailing list, but we won't target blueprints to those teams specifically. We can make the teams official again once they have some activity and a clear person to be responsible for running them. Please keep in mind that I'm not saying that this stuff isn't important, just that there is no need to separate out if there isn't a lot of activity. Specifically the teams which haven't really been very active are: 1. Nova Upgrades Team http://wiki.openstack.org/Teams#Nova_Upgrades_Team 2. Nova Auth Team http://wiki.openstack.org/Teams#Nova_Auth_Team 3. Nova Security Improvements Team http://wiki.openstack.org/Teams#Nova_Security_Improvements_Team Agreed, so far it's mostly been individual efforts, nothing coordinated. 4. Nova Operational Support Team http://wiki.openstack.org/Teams#Nova_Operational_Support_Team 5. Nova EC2 API Team http://wiki.openstack.org/Teams#Nova_EC2_API_Team Im going to leave the ec2 team for now, because it is relatively new. If anyone feels that the other teams above should not be folded back in, please let me know. The EC2 team has been meeting a few times already, and started triaging and working on EC2 bugs. I think it belongs to the active category :) Regards, -- Thierry Carrez (ttx) Release Manager, OpenStack ___ 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] Openstack with XenServer
Hi,all, I am trying to install OpenStack with XenServer.I followed what this page says,but with no luck. (http://wiki.openstack.org/XenServerDevelopment) Is there any one who successfully installed this? Is there any guides available beyond the page above? What's more, I tried this way: https://github.com/cloudbuilders/devstack/tree/master/tools/xen and it doesn't seem to work either.Any thoughts? Thanks very much. Yours, Jeff ___ 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] Providing packages for stable releases of OpenStack
On 06 Dec 2011 - 13:52, Duncan McGreggor wrote: Yikes! I forgot an incredibly important one: * What is the migration path story (diablo to essex, essex to f, etc.)? I think it was going to be the Upgrades Team? ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
Can you talk a little more about how you want to apply this failure notification? That is, what is the case where you are going to use the information that an operation failed? In my head I have an idea of getting code simplicity dividends from an everything succeeds approach to some of our operations. But it might not really apply to the case you're working on. Sandy Walsh sandy.wa...@rackspace.com said: For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
Sure, the problem I'm immediately facing is reclaiming resources from the Capacity table when something fails. (we claim them immediately in the scheduler when the host is selected to lessen the latency). The other situation is Orchestration needs it for retries, rescheduling, rollbacks and cross-service timeouts. I think it's needed core functionality. I like Fail-Fast for the same reasons, but it can get in the way. -S From: openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net [openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net] on behalf of Mark Washenberger [mark.washenber...@rackspace.com] Sent: Wednesday, December 07, 2011 11:53 AM To: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit Can you talk a little more about how you want to apply this failure notification? That is, what is the case where you are going to use the information that an operation failed? In my head I have an idea of getting code simplicity dividends from an everything succeeds approach to some of our operations. But it might not really apply to the case you're working on. ___ 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] [QA] Team meeting at 12 EST today
Hey all, A quick reminder that the QA team has our weekly meeting on #openstack-meeting in about 30 minutes. 12:00 EST 09:00 PST 17:00 UTC See you there, -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] [Orchestration] Handling error events ... explicit vs. implicit
Gotcha. So the way this might work is, for example, when a run_instance fails on compute node, it would publish a run_instance for uuid=blah failed event. There would be a subscriber associated with the scheduler listening for such events--when it receives one it would go check the capacity table and update it to reflect the failure. Does that sound about right? Sandy Walsh sandy.wa...@rackspace.com said: Sure, the problem I'm immediately facing is reclaiming resources from the Capacity table when something fails. (we claim them immediately in the scheduler when the host is selected to lessen the latency). The other situation is Orchestration needs it for retries, rescheduling, rollbacks and cross-service timeouts. I think it's needed core functionality. I like Fail-Fast for the same reasons, but it can get in the way. -S From: openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net [openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net] on behalf of Mark Washenberger [mark.washenber...@rackspace.com] Sent: Wednesday, December 07, 2011 11:53 AM To: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit Can you talk a little more about how you want to apply this failure notification? That is, what is the case where you are going to use the information that an operation failed? In my head I have an idea of getting code simplicity dividends from an everything succeeds approach to some of our operations. But it might not really apply to the case you're working on. ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
Exactly! ... or it could be handled in the notifier itself. From: openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net [openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net] on behalf of Mark Washenberger [mark.washenber...@rackspace.com] Sent: Wednesday, December 07, 2011 12:36 PM To: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit Gotcha. So the way this might work is, for example, when a run_instance fails on compute node, it would publish a run_instance for uuid=blah failed event. There would be a subscriber associated with the scheduler listening for such events--when it receives one it would go check the capacity table and update it to reflect the failure. Does that sound about right? ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
Hi Sandy, I'm wondering if it is possible to change the scheduler's rpc cast to rpc call. This way the exceptions should be magically propagated back to the scheduler, right? Naturally the scheduler can find another node to retry or decide to give up and report failure. If we need to provision many instances, we can spawn a few green threads for that. Yun On Wed, Dec 7, 2011 at 10:26 AM, Sandy Walsh sandy.wa...@rackspace.com wrote: For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
True ... this idea has come up before (and is still being kicked around). My biggest concern is what happens if that scheduler dies? We need a mechanism that can live outside of a single scheduler service. The more of these long-running processes we leave in a service the greater the impact when something fails. Shouldn't we let the queue provide the resiliency and not depend on the worker staying alive? Personally I'm not a fan of removing our synchronous nature. From: Yun Mao [yun...@gmail.com] Sent: Wednesday, December 07, 2011 1:03 PM To: Sandy Walsh Cc: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit Hi Sandy, I'm wondering if it is possible to change the scheduler's rpc cast to rpc call. This way the exceptions should be magically propagated back to the scheduler, right? Naturally the scheduler can find another node to retry or decide to give up and report failure. If we need to provision many instances, we can spawn a few green threads for that. Yun On Wed, Dec 7, 2011 at 10:26 AM, Sandy Walsh sandy.wa...@rackspace.com wrote: For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
*removing our Asynchronous nature. (heh, such a key point to typo on) From: openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net [openstack-bounces+sandy.walsh=rackspace@lists.launchpad.net] on behalf of Sandy Walsh [sandy.wa...@rackspace.com] Sent: Wednesday, December 07, 2011 1:55 PM To: Yun Mao Cc: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit True ... this idea has come up before (and is still being kicked around). My biggest concern is what happens if that scheduler dies? We need a mechanism that can live outside of a single scheduler service. The more of these long-running processes we leave in a service the greater the impact when something fails. Shouldn't we let the queue provide the resiliency and not depend on the worker staying alive? Personally I'm not a fan of removing our synchronous nature. From: Yun Mao [yun...@gmail.com] Sent: Wednesday, December 07, 2011 1:03 PM To: Sandy Walsh Cc: openstack@lists.launchpad.net Subject: Re: [Openstack] [Orchestration] Handling error events ... explicit vs. implicit Hi Sandy, I'm wondering if it is possible to change the scheduler's rpc cast to rpc call. This way the exceptions should be magically propagated back to the scheduler, right? Naturally the scheduler can find another node to retry or decide to give up and report failure. If we need to provision many instances, we can spawn a few green threads for that. Yun On Wed, Dec 7, 2011 at 10:26 AM, Sandy Walsh sandy.wa...@rackspace.com wrote: For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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
Re: [Openstack] Openstack with XenServer
Hi Jeff, Can you be more specific about what doesn't work? There are lots of people using OpenStack with XenServer, including Citrix and Rackspace, so I can guarantee that it works! The docs are lacking though, that's for certain. Where did you get stuck? Thanks, Ewan. From: openstack-bounces+ewan.mellor=citrix@lists.launchpad.net [mailto:openstack-bounces+ewan.mellor=citrix@lists.launchpad.net] On Behalf Of jeffrey coho Sent: Wednesday, December 07, 2011 4:32 AM To: openstack@lists.launchpad.net Subject: [Openstack] Openstack with XenServer Hi,all, I am trying to install OpenStack with XenServer.I followed what this page says,but with no luck. (http://wiki.openstack.org/XenServerDevelopment) Is there any one who successfully installed this? Is there any guides available beyond the page above? What's more, I tried this way: https://github.com/cloudbuilders/devstack/tree/master/tools/xen and it doesn't seem to work either.Any thoughts? Thanks very much. Yours, Jeff ___ 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] Providing packages for stable releases of OpenStack
On Tue, 2011-12-06 at 23:56 +0100, Loic Dachary wrote: I think there is an opportunity to leverage the momentum that is growing in each distribution by creating an openstack team for them to meet. Maybe Stefano Maffulli has an idea about how to go in this direction. The IRC channel was a great idea and it could become more. It seems that there is general consensus that OpenStack should be the good upstream provider and ease the job of downstream distributions. There is an informal team that is coordinating the packaging effort. Would it help if this team became more formal? What would it mean for the openstack-packaging team to be 'formally' recognized? /stef ___ 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] [Orchestration] Handling error events ... explicit vs. implicit
On Wed, Dec 7, 2011 at 7:26 AM, Sandy Walsh sandy.wa...@rackspace.com wrote: For orchestration (and now the scheduler improvements) we need to know when an operation fails ... and specifically, which resource was involved. In the majority of the cases it's an instance_uuid we're looking for, but it could be a security group id or a reservation id. With most of the compute.manager calls the resource id is the third parameter in the call (after self context), but there are some oddities. And sometimes we need to know the additional parameters (like a migration id related to an instance uuid). So simply enforcing parameter orders may be insufficient and impossible to enforce programmatically. A little background: In nova, exceptions are generally handled in the RPC or middleware layers as a logged event and life goes on. In an attempt to tie this into the notification system, a while ago I added stuff to the wrap_exception decorator. I'm sure you've seen this nightmare scattered around the code: @exception.wrap_exception(notifier=notifier, publisher_id=publisher_id()) What started as a simple decorator now takes parameters and the code has become nasty. But it works ... no matter where the exception was generated, the notifier gets: * compute.host_id * method name * and whatever arguments the method takes. So, we know what operation failed and the host it failed on, but someone needs to crack the argument nut to get the goodies. It's a fragile coupling from publisher to receiver. I'm just wondering if we can get the notification message down to something more standardized, and avoid including the full argument list. That is one way to reduce the coupling. What is the minimum information we need to know when a failure occurs ? I think we have operation host it failed on, instance_id, migration_id (maybe) reservation_id, (maybe) security group id (maybe) If we can avoid cracking open the remaining arguments, a list this long might be manageable. One, less fragile, alternative is to put a try/except block inside every top-level nova.compute.manager method and send meaningful exceptions right from the source. More fidelity, but messier code. Although explicit is better than implicit keeps ringing in my head. I like explicit better than implicit, but I think we need to trigger off any and all exceptions to make all of this reliable. Or, we make a general event parser that anyone can use ... but again, the link between the actual method and the parser is fragile. The developers have to remember to update both. Opinions? -S ___ 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] Providing packages for stable releases of OpenStack
On 07 Dec 2011 - 08:22, Mark McLoughlin wrote: On Tue, 2011-12-06 at 10:11 -0800, Duncan McGreggor wrote: On 06 Dec 2011 - 14:28, Thierry Carrez wrote: So the general consensus so far on this discussion seems to be: (0) The 2011.3 release PPA bears false expectations and should be removed now. In the future, we should not provide such PPAs: 0-day packages for the release should be available from the last milestone PPA anyway. (1) OpenStack, as an upstream project, should focus on development rather than on providing a production-ready distribution. (2) We could provide daily builds from the stable/diablo branch for a variety of releases (much like what we do for the master branch), but those should be clearly marked not for production use and be best-effort only (like our master branch builds). (3) This should not prevent a group in the community from working on a project providing an openstack on Lucid production-ready distribution if they so wishes. This project would just be another distribution of OpenStack. This doesn't seem like enough to me. OpenStack isn't just a library; it's a fairly substantial collection of software and services, intended to be used as a product. If it can't be used as a product, what's the use? Someone made the suggestion that a new OpenStack group be started, one whose focus is on producing a production-ready, distribution-ready, release of the software. So can we add one more (need some help with wording, here...): This paragraph makes it sound like you think OpenStack upstream should produce production-ready binary packages? Nope, just production-ready code. Doing the many things necessary so that: 1) when someone wants to build a binary package, they have everything that they need to do so; 2) when the OpenStack components are installed from these packages, they run well; 3) when someone wants to read the documentation for that release, it's available, up-to-date, and accurate; 4) when the unit tests are run for a given release, they all pass on a properly configured system (which is documented, supporting #3 above); 5) when the next release is out, upgrading is a clear and straight-forward process. The idea is not that these are *all* missing; rather, it appears to me (and others on this list) that the activities outlined above are not well-coordinated and/or prioritized. (4) OpenStack will accept and foster a new project, one that is not focused on development, but rather the distribution and it's general stability. This distro project will be responsible for advocating on behalf of various operating systems/distros/sponsoring vendors for bugs that affect performance and stability of OpenStack, or prevent an operating system from running OpenStack. This sounds like you want want OpenStack to focus more on the quality of its releases. Agreed, but it's not just about QA -- there is a lot involved here. With lots of different teams. My thought is that if there was an actual point of contact (person or team) that was responsible for coordinating on the specific set of areas that were agreed to have the most impact on creating a production-ready product, operators/users, vendors, etc., would have much more confidence in OpenStack as a dependable platform that business-critical systems could put their trust in. It also sounds like you want more co-ordination upstream between downstream distributions of OpenStack. I think there's already pretty good co-ordination. Is there any specific problem you've seen on that front? I think that if we had something in place like the 5 numbered points above (and other points that have been mentioned in this thread), such that an operator could quickly and easily get from 0 to production-ready in a few short steps, OpenStack would be truly unstoppable. And an obvious choice for both those with no budget and those with big budgets. d ___ 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] Providing packages for stable releases of OpenStack
On 12/07/2011 10:32 PM, Stefano Maffulli wrote: On Tue, 2011-12-06 at 23:56 +0100, Loic Dachary wrote: I think there is an opportunity to leverage the momentum that is growing in each distribution by creating an openstack team for them to meet. Maybe Stefano Maffulli has an idea about how to go in this direction. The IRC channel was a great idea and it could become more. It seems that there is general consensus that OpenStack should be the good upstream provider and ease the job of downstream distributions. There is an informal team that is coordinating the packaging effort. Would it help if this team became more formal? What would it mean for the openstack-packaging team to be 'formally' recognized? Thierry Carrez and Mark McLoughlin essentially asked the same question on IRC a few hours ago. You are on the same line ;-) I honestly don't know how to express myself more clearly than I did in my previous mails. The Debian GNU/Linux packaging team is doing fine. Since you are all comfortable with the fact that it operates independently, there probably is no need to create tighter relationships. If you ever feel that this needs to be discussed again, let me know. Cheers attachment: loic.vcf signature.asc Description: OpenPGP digital signature ___ 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] Delete server spec
Hi folks I wanna make Delete server spec clear. The API doc says, When a server is deleted, all images created from that server are also removed http://docs.openstack.org/api/openstack-compute/1.1/content/Delete_Server-d1e2883.html IMO, all images is vm images which stored on compute node instances_path. Is this correct? Cheers Nati ___ 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] Delete server spec
I would interpret that to include the snapshots - but I'm not sure that is what I'd expect as a user. On Wed, Dec 7, 2011 at 5:05 PM, Nachi Ueno ueno.na...@nttdata-agilenet.com wrote: Hi folks I wanna make Delete server spec clear. The API doc says, When a server is deleted, all images created from that server are also removed http://docs.openstack.org/api/openstack-compute/1.1/content/Delete_Server-d1e2883.html IMO, all images is vm images which stored on compute node instances_path. Is this correct? Cheers Nati ___ 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] Delete server spec
Hi Jessy Thanks. Hmm, there are no implementation of cleanup snapshot images. IMO, Snapshot image should not deleted, in case of API request mistake. https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L540 2011/12/7 Jesse Andrews anotherje...@gmail.com: I would interpret that to include the snapshots - but I'm not sure that is what I'd expect as a user. On Wed, Dec 7, 2011 at 5:05 PM, Nachi Ueno ueno.na...@nttdata-agilenet.com wrote: Hi folks I wanna make Delete server spec clear. The API doc says, When a server is deleted, all images created from that server are also removed http://docs.openstack.org/api/openstack-compute/1.1/content/Delete_Server-d1e2883.html IMO, all images is vm images which stored on compute node instances_path. Is this correct? Cheers Nati ___ 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] Delete server spec
I would agree with that. If I delete a server instance, I don't want to destroy snapshot images I took of that server... -jay On Wed, Dec 7, 2011 at 8:50 PM, Nachi Ueno ueno.na...@nttdata-agilenet.com wrote: Hi Jessy Thanks. Hmm, there are no implementation of cleanup snapshot images. IMO, Snapshot image should not deleted, in case of API request mistake. https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L540 2011/12/7 Jesse Andrews anotherje...@gmail.com: I would interpret that to include the snapshots - but I'm not sure that is what I'd expect as a user. On Wed, Dec 7, 2011 at 5:05 PM, Nachi Ueno ueno.na...@nttdata-agilenet.com wrote: Hi folks I wanna make Delete server spec clear. The API doc says, When a server is deleted, all images created from that server are also removed http://docs.openstack.org/api/openstack-compute/1.1/content/Delete_Server-d1e2883.html IMO, all images is vm images which stored on compute node instances_path. Is this correct? Cheers Nati ___ 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
Re: [Openstack] Nova Subteam Changes
excellent ideas. I especially like the standardized list of headers. just to be sure, mondays at 2100utc? if so, no conflicts on my end On Wed, Dec 7, 2011 at 4:51 AM, Thierry Carrez thie...@openstack.orgwrote: Vishvananda Ishaya wrote: 2) *Closing down the team mailinglists.* Some of the lists have been a bit active, but I think the approach that soren has been using of sending messages to the regular list with a team [header] is a better approach. Examples: [db] Should we use zookeeper? [scaling] Plans for bursting I realize that this will lead to a little more noise on some of the channels, but I think it makes sure that we don't isolate knowledge too much Could we have a standardized list of headers ? Those would double as tags for Nova bugs, so the scaling team could see a list of bugs in their area by searching for the scaling tag. 3) *Closing teams.* Some of the teams haven't really started having much activity. I'm going to move these teams into a WANTED section on the teams page here: http://wiki.openstack.org/Teams For now, people can just discuss stuff on the main mailing list, but we won't target blueprints to those teams specifically. We can make the teams official again once they have some activity and a clear person to be responsible for running them. Please keep in mind that I'm not saying that this stuff isn't important, just that there is no need to separate out if there isn't a lot of activity. Specifically the teams which haven't really been very active are: 1. Nova Upgrades Team http://wiki.openstack.org/Teams#Nova_Upgrades_Team 2. Nova Auth Team http://wiki.openstack.org/Teams#Nova_Auth_Team 3. Nova Security Improvements Team http://wiki.openstack.org/Teams#Nova_Security_Improvements_Team Agreed, so far it's mostly been individual efforts, nothing coordinated. 4. Nova Operational Support Team http://wiki.openstack.org/Teams#Nova_Operational_Support_Team 5. Nova EC2 API Team http://wiki.openstack.org/Teams#Nova_EC2_API_Team Im going to leave the ec2 team for now, because it is relatively new. If anyone feels that the other teams above should not be folded back in, please let me know. The EC2 team has been meeting a few times already, and started triaging and working on EC2 bugs. I think it belongs to the active category :) Regards, -- Thierry Carrez (ttx) Release Manager, OpenStack ___ 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