 I'm David Duncan. I'm part of the Fedora 35, Fedora Cloud working group and I wanted to talk to you today about today here in November not in August about Fedora 35 and tell you why you should use 30 Fedora in the cloud. So we're part of the focus of our special interest group is building the images and we consider that our responsibility to do that for any cloud that we possibly can. So we want to make sure that there are images available pretty much everywhere. Some of the images that we we work on today are the GCP cloud images, the vagrant images and then obviously the Amazon cloud images and then we have a base cloud image that can be used in many different places open stack different types of virtualization and then we have just a standard compressed raw that you can take and do many things with by converting it through the various tools that are available. We're working on multiple architectures. We support x8664 and art64 directly and then S390 we do with with some friends and and that has been an interesting thing for Fedora 35 because we've made some changes. We have a bi-weekly meeting on Thursdays, 1400 UTC and the cloud SIG, Pager has has a list of people who want to be notified if you want to be notified hang around in the pound Fedora cloud channel on matrix or IRC or participate in the cloud list or both hopefully and add your name and we'll make sure that we let you know when the meetings are starting every time the the group and the directors of this group do not force do not include being an expert at some cloud technology. The requirements for participating are an interest in participating so if you are a beginner or advanced user that doesn't matter to us we would love to have you there we'd love to have your opinion and we'd love to make sure that if we're working on something that is a part of your your goals or your experience that you have an opportunity to participate so please join us anytime that you feel you know any fine anytime you feel like you have an opportunity so we had a conversation not too long ago and some metrics appeared in that conversation from Matthew to show a little bit about how we're doing and and thank you Matthew for putting these together and and keeping keeping such great metrics and to giving us something that we could rely on to tell us where we are in this in this process so thanks Matthew but looking at what are considered to be persistent which Matthew said our instances that are up for and up and reporting for more than one week the cloud is roughly 15.2% of the of the the cloud variants that are being used for long term and then if you move as we move into the ephemeral those that don't report after one week this is you see that it makes up a considerably larger almost double the number of instances that are ephemeral which is expected right cloud is not meant usually cloud is meant to be a fail-only architecture or up for a specific task and then down and that is that's wholly expected that we would see a much larger part of that part of this going to the what we call the ephemeral use and on our out this is not just I mean this is the cloud images but where they're used we're not I don't think we're tracking that at all like we're not looking at whether or not this comes in from a specific public cloud or a specific location we're looking at what what the what the designation is in the release okay so let's talk a little bit about what shipped in and fedora 35 I'm pretty excited about this hybrid boot so one of the things that I think we we look we look at on the on the cloud side is the emergence of the UEFI support and and that has not been as pervasive as we as you might expect a lot of the UEFI boot process that was previously used for things like windows was not was emulated it wasn't in fact actually UEFI and this gives us the the the boot parameters or the in the boot partitions that are required to do both so we can do both the UEFI boots we can do classic boot but classic BIOS boot and and we're ready to do new things for example secure boot in the in the cloud images when that becomes available or if that becomes something that our users are looking for and big moves we followed the trend that we you know is in work station here by moving to butterfs by default this adds a lot of capabilities that we didn't have previously and I think this is going to be very interesting I'm looking forward to doing some solutions based articles over the over the coming weeks to show what what this really means because this gives us some some ability to do some bind mount points and take control of specific component parts of the or specific parts of the disk for removing execution bits and things like that that we didn't have before but work that are typically expected in things like CIS hardening or Stig hardening things like that don't don't allow for an exception around the partitions and since a cloud volume doesn't have partitions it is almost by definition an exception on every case on every file system level case so this gives us some some breathing room around those those requirements and makes it possible for us to be more consistent with them also one of the considerations that we had you know if you're looking at server server has support for very large arrays of disks and we don't necessarily have that in cloud I mean you can you can acquire a large volume of storage but for the most part the storage redundancy is handled behind the scenes by the hyperscalers and so we expect the we expect that using better FS on top even though it doesn't have support for things like RAID 5 or RAID 6 isn't really a big problem because those are those RAID types are being used behind the scenes in the hyperscalers storage but you can look at the whole configuration and and see that you know this is something we've done and then in order to get this done we had to push through several changes around BIOS we had to make some modifications in Anaconda there are some other considerations that we have for for next generation that I think will be you know that as they evolve will be exciting for us like like full support for butterfs in the in the boot sector that'll that'll carry us over into a full fully butterfs configuration and we made some changes in the cloud images for vague for vagrant files one of the things that we found was that there was a bug that was increasing the size of the the volumes for the vagrant images to greater than 40 gig which obviously nobody wants to download a 40 gig vagrant file that's pretty much antithetical to what needs to happen but Chris Murphy and and Neil Gampa actually took care of creating a better trim experience in our in our kickstart configurations and zero that out so that it would it would be appropriately resized during the vagrant configuration so there's been some interesting developments here over the course of the Fedora 35 release and we are really we had to stop sort of stand back and take a really good look at where we were and what we were working on and whether or not we thought that was a this was the right way you know right direction for us to go to create cloud based images and we've made a decision as of Thursday in fact we had a vote for the for throughout the whole cloud working group to determine that we want to reestablish cloud based images as an addition so that we can participate more in the website revamp and be more formally associated with the addition status so we're working right now on updating the product requirements so that our PRD is current and can be ready for review but we also consider this to be a real responsibility so we're not we're not looking at what it is that just needs to be done to be you know to have addition status we think that the responsibility goes much farther into into the community and so we want to establish a pipeline for ensuring that we're participating in the marketing and that we're participating in blog posts and also increase the number of testing and QA contributions so I'll refer you back to the first slide I put up here where I was saying that you know we're looking forward to having your participation there's a lot more things to do here obviously than just code on on cloud images so if you're looking at trying to make you know create a scenario where you're working with other tools or utilities or or different kinds of technology then we're excited to hear from you if you're looking at how you can help with the QA we have a ton of things that need to be done plumbing for tests in each one of the individual clouds we've got lots of things that need to be done here that we would love to have help with and something that's near and dear to my heart and I keep pushing this and I will continue is we really want to see this develop into something where we're building solutions and customers are users really not not customers but users want us to want to to develop they want to have understanding of something kind of end to end and we want to bring as many people as we can into qualifying and and seeing and looking at how those solutions work so that so that people who are doing research don't have to question why they would want to use fedora in their in their solution requirements if they have some study to do we want to make sure that they understand that they can get all of the tools that they want inside of fedora edition similar to the way that the neuro fedora team works on their span and to look at ways that we can collaborate and and curate those those sets of solution configurations for for people to use in in the future and that means more articles more writing better collaboration with teams that we are not really working with as much as I think we should especially os build I really feel like os build is a is a wave of the future and and then there are some really fun things going on in cockpit now even preliminary support for butter fs in the in the storage management and really look forward to putting a lot of effort and time into making those work together and then recently we've been chatting back and forth with some of the members of the server group to talk about creating images more consistent with their sick goals and I think that that's a really important effort to be you know effectively supporting other special interest groups for the work that they're doing and I think we can do that with workstation too in the context of virtual desktop interfaces so there are a lot of things that we need to focus on that are interesting to providers like cloud providers and some of the things that they run in some of the issues that they run into are the things that are trend setting for them and we want to make sure that when they look for how they set that bar our goal is to ensure that regardless of what that of what that infrastructure is that they look to fedora as a leader in how they evaluate and identify themselves as forward as a as a forward leader in that in their space some of the things that they look to and some of the things that we're trying to focus on are like faster boot times and then we also want to have a faster model for continuous integration and testing if we have more detail back from their our providers that makes it back into fedora um just generally especially in the qa process the uh the information that information can feedback into uh into our development all the way down from the kernel and we can run you know we can run into problems long before the the public provider or the providers make them public and uh we think that that's a really important part is of our mission is to get in front of a lot of those issues that would show up after uh you know after the fact after after the providers make them public after the providers announce them um and we want to make sure that we're uh we're giving them enough information and we're getting enough feedback from them uh that we can we can get that we can make changes even before we know what they're going to do um some of the other things that we've thought about are that we have a similar mission for the container configurations and we think that might be a great fit for the cloud sig umbrella we know that and i know that those are well i mean everything in fedora is maintained by the people who have the motivation to to um uh to uh to take on the projects but uh you know it's always been one of my favorite things that we don't have a um you know we don't have package owners we have packages and those packages do what's necessary to take on take on that space um take on the space that they want they want to deliver in and i think that that's a great experience um yeah i think with that let's just move on to uh questions and comments discussions i see the block of butter question has already been answered i don't think i have to uh i don't think i have to answer that one comments considerations that scare anybody well listen you know y'all know where to find me i'll be around and uh if you have any other people who who want to um uh who if you know anyone that you think should be involved you know give them a nudge let us know that they're coming and uh we'll we'll make a place for them um we have a lot of fun working on this project and uh look forward to working on it for a long time to come thanks i'm going to hang around in case somebody does have a question thanks everybody for attending it's been great yawn i think i think biweekly is enough right now i think we have a lot of work to do and and there's a lot of chat you know we we have a lot of chatter in the in the in the um in the matrix group um if there's something uh oh there was something i didn't mention actually and that is that we really want to curate the tools um that are associated with the cloud cloud experience like the azure cli oh okay thanks Matthew um i think uh the thing that we are going to put a whole lot of time and energy into for the next year uh is going to be first the uefi space and then uh i think we're probably i think what i'm looking most forward to is finding a way to to build uh secure boot um for our for our images so that customers you know users who are doing who are trying to do something like a secure boot are going to do that and then the other thing is the trusted execution environment i really want to um integrate the ability to run trusted code in uh in the trusted execution environment that intel created um directly and to have a process for building the images for that trusted execution environment in our in our process so that we know exactly what's going into that blob end to end um each one of the the trusted execution environment i think is is one that we're we're going to try and focus on and then tools the the azure cli the aws cli um you know the the uh google api tools all of those are things that we're we're focusing in on really heavily and i've got to say a big thanks big shout out to major hayden uh for for uh building uh i think the right number is a gazillion uh dependencies for uh for the the azure cli and we're working on the same thing together for the aws cli where uh there are a gazillion dependencies none of them make sense so we're we're we're working on that uh together and uh trying to make sure that those go through our the the fedora build process the way that we expect so those are the things though those curations i think are the things that i'm most excited about um moving from from 35 to 36 and beyond and then um uh doing the work to become uh become a real addition uh once again i think is is also heavy on our on our plate and with that i think we're we're closing time thanks everybody for for participating and uh see you in the next tap