 Hi. This is Nikhil Kumar. I'm the PTA for OpenStack Images Service. It's more popularly known as Glance. The purpose of these slides is to share with you some details about the developments happened in Juneau and those that are planned for Kilo. Before we start off, I would like to thank you all for taking the time to listen to this presentation. Hope you find it useful. Next slide, please. So what is OpenStack Glance? It's a set of services which are used to store, browse, share, distribute, and manage bootable disk images within the stack of OpenStack. What that means is, when you attempt to boot an instance, the OpenStack compute service makes a call to Glance. It will fetch the metadata as well as the image data. Glance has a set of drivers for the backend storage system which will enable it to fetch the image data in consideration. This way, it has the ability to keep the image data's consistency in the check as well. So images and safe hands. Next slide, please. Okay, let's have a flashback to Juneau. These are the major changes that have happened in the previous cycle. So asynchronous processing was a major change that was added. It has added a new API for performing asynchronous tasks on the image service. There are three kinds of tasks that can be performed by the user, that is import, export, and cloning. Currently, import task has been implemented and is supported to a certain extent. We are planning to add the export and cloning support in the kilo cycle. Basically, these tasks will let you perform some kinds of operations on the image data as well as the image metadata. Another change that has happened in this major OpenStack Glance is separation of the drivers into a separate library called Glance Store. It's still under the same program. However, the separation will help with the maintenance as well as enhancement for the existing drivers. It will also help add new backend drivers without much impedance that would have been possible in the Glance Service. Another change that we are really proud of is the metadata definition scatter. It's a new addition to this program. It will help operators to store rich information associated with the metadata. The restricted policy for downloading images is basically an enhanced capability for operators to control the download policies at a much granular level. Enhanced server service is basically a Daemon process which is run along with the Glance API. It is an alternate approach for deleting images and has been designed to better serve the performance for image collision. Besides that, there are some performance improvements that are constantly happening to the service. So please be on look out for that as well. Next slide, please. So there has been a theme change in Glance in the Tuner Cycle. It has become a platform for storing wider set of data assets such as images for NOVA and templates for HEAT. Next slide, please. Another great addition to Glance Service has been the concept of specifications. Basically, specifications is a process as well as a platform and it will help foster better communication as well as provide a coherent platform for both operators as well as developers. It will help discuss the design specifications from different dimensions of the project. And we hope that it will help individuals get acquainted and add her to the open stack as well as Glance standards. You can think of specifications as a formal statement for describing the proposed changes to the code base. This way, there is a clear description of what the developer or the operator is thinking about making a change within the code base. You can find the respective links for proposed as well as approved specifications in the slides. Next slide, please. So we think of Glance in two ways. One is the features, another aspect being pseudo features. So all of us are quite familiar with the new and cool things that are added to the services quite often. And those are features they can be introduced via bugs, but more often they are introduced via specifications. We encourage you to go via the router specifications. However, sometimes it might just be convenient as well as less of a time hassle if a bug has been filed against the proposed change. Of course, the discussions are welcome on all fronts. So please be involved in the same. Another aspect of the project being pseudo features, we focus on stability. We also focus on releases of the different libraries. Currently, we have Python Glance Client as well as the Glance Total Libraries within the program. It is also a way of thinking for enhancing and improving performance or content study issues. If you have any suggestions or concerns regarding the same, please feel free to reach out to liaisons of the respective aspects of pseudo features. You can find the IRC, IRCnext as well as the names of individuals in the link below. Next slide, please. Okay, so some of the new features that we are planning to add in Kilo are listed in this slide. One of the important things that we have done is bumped up the version to 2.3. It's a minor version bump. There was a small change that happened within the API and we wanted the API to be compatible for the operators. Another really important concern for us is the stability of the Glance Total Library. It was separated out late in the cycle in Chuno. It's still in the alpha phase of the development. We would like to improve the release management of the same as well as provide a cleaner API. We hope that will help other services within the OpenStack to be able to use it and leverage the store drivers without duplicating the code themselves. After the incorporation of the metadata definitions catalog, we felt the need to include a better way for operators to search through them. So we have been constantly attempting to improve this service as well as provide a new way of searching either the image data or the metadata definitions. That is being planned by the catalog indexing service. The proposal is still in progress. There has been some work that has been proposed via code as well as specifications. You are welcome to provide your feedback on either of those. There is another proposal after the theme change and that is of artifacts. Artifacts being the generic data asset concept. It will help add some cool new features to the existing domain of image objects that are currently designed in Glance. As well as it will fight in the scope for storing different kinds of data assets as per the need of specific projects. Again, it's still a proposal. We are still designing and finalizing some of the final details. There is a plan for enhancing the asynchronous processing and that is being planned by the Taskflow service which has been developed hand in hand with other open-sack services. The first big thing that Taskflow service would help us achieve is the concept of image conversion and this image conversion concept comes as a side effect of the import process within the image service. The import process being deployers opening up the grounds for users to upload their custom image data and that can be done by the scripts. Different checks can be performed on the same. However, if the user wants to upload a specific kind of disk image which is not supported within the cloud then image conversion concept really helps to transfer or transpose either or of these operations on the disk images and store it within the Glance being compatible with the infrastructure service provision. Besides that, we have been constantly attempting to improve the scrubber service for performance as well as future additions to the image deletion process. Functional API has been proposed for images to perform reactivation and deactivation of images. Again, this is a side product of the import process if the deployers or the admin status cannot find a custom image to be compatible or safe, they can leverage this functional API and either reactivate the image or reactivate the image after certain checks. Community image sharing is an enhancement to the existing concept of image sharing. It will just enable the users to basically use the custom images or the deployer images and share it with the wider audience. Next slide, please. Some of the miscellaneous features that we are attempting to add in Kilo is to help the operators as well as admins use error codes and identify bugs and issues within the code in a smoother fashion. It will hopefully, as well, enhance the consistency checks between different OpenStack services. Hopefully, it will provide a standard, some discussion going around behind scenes on a wider OpenStack platform, so we will attempt to add her and adopt those within Glance as well. A recent addition for separation of logging for Glance Manage should help the operators to identify any issues that they see just within the specific Glance Manage operations most common of them being the database migrations. Multiple containers for single-dent search backend is an enhancement within the search driver. It should help improve the performance in certain kinds of deployments and it's specific to the search backend. We keep the platform open for enhancing the drivers and this is just one good example of being the same. We are also planning to help NOAA or the OpenStack Compute Service use the version 2 of Glance API and there are some enhancements that need to be done either to the Glance client or the Glance service and Glance version 2.3 was one of them. More work is being done in Clilo and we hope to continue the same. Enhancement for adoption of Glance Store outside of Glance is another undertaking that is the main objective of the same being to provide a cleaner store API for other OpenStack services as well as users to leverage the Glance Store drivers and use the backend in a more effective manner. Next slide please. So those were some of the details I wanted to share with you all about the Clilo as well as the JunoCycle. However, if you have any more questions or suggestions or concerns I'm always happy to chat with you. Some of the medium of communication can be IRC in my opinion. However, I personally prefer weekly meetings because a complete set of or majority of the Glance developers who are actively providing good feedback are available there and it just helps everyone to get in sync with what's happening, what needs to happen, come to and go ahead and design. So please feel free to reach out to me preferably on the weekly meeting and if you have any more questions please send me the same. Thank you for listening to this presentation. I hope you find it useful. Have a nice day. Bye.