Building an OKD 4 Home Lab with special guest Craig Robinson

[Music] you good morning good evening good afternoon wherever you are hello welcome thank you for joining us today on openshift TV we are very very happy to have you here I am joined by our first ever customer on the streams this is really like a momentous occasion for me personally to have someone out there that is willing and able to join the stream that is a customer of ours so I'm super happy to be joined by Craig Robinson the easy site reliability engineer at East Carolina University if you're a pirate shout out to y'all I will let him introduce himself Craig please take it away sure yeah so I'm Craig Robinson I worked for ECU we're based in Greenville North Carolina in February this year I created a guide on installing okd 4.3 in a home lab and okd is the community version openshift I created the guide because at the time there wasn't anything for okd out there so it wasn't specific to okay d so since then I'm creating a guide on version 4.3 and now for point 5 which I finished up yesterday so basically what I want to do is just run through this guide and in this guide yeah yeah like I think this is important for folks because we were talking beforehand right like everybody right now is kind of stuck at home and we all like I have two nooks some raspberry PI's we all have some some hardware laying around so getting okd up and running in a home lab is I think a great thing right now so please take it away your your your blog posts have been fantastic over the the months so I'm sure you've got this well in hand okay yeah thanks I did use the sxi in this in the God without vsphere on purpose I wrote it so it's kind of universal so you can adapt it to your own platform I've had a few folks who have had look with proxmox and libvirt and even somebody who tried it in a digital ocean droplet not sure if that got working or not but he was I tried to help him with it so anyway I'm just gonna jump right in and go through the guide so quick intro I talked about hoodie kay and Fedora core OS and you know my home lab was relatively inexpensive less than five hundred dollars and I ended up getting a an older day little service got 96 gigs of ram so here's a quick overview of our virtual machines that we're going to be using there's a okt for bootstrap 3 control planes and to compute nodes and they're all going to run the Fedora core OS and they'll have four virtual CPU 16 gigs of ram and 120 gigs storage and those are the IP addresses there's the okd for services machine that's actually what I'm logged into now and in what you're looking at and it's going to run our DNS or load balancer web services and NFS and it's running CentOS a guy in finally we want to have a pfsense virtual machine that we're gonna use for the router and load balancer or I'm sorry the router and DCP reservations so here's a network diagram this is a my home network and these are the different virtual machines to the the pfsense VM is going to have a wayne interface which is connected to my home network and the lan interface is gonna be a 1.1 and the okd for services machine i added an extra interface so i can get to it through RTP and this is how you create the network and being worse but I created a different network and I'll show you how to do that in VMware so if you go to your web console and then click on networking and had a port group i called mono KD which is already used and I used VLAN 20 because it wasn't in used but that way it's segmented from your home network if you're in a home live environment so that's that part of the guide next I created the okd for services VM I'm not going to show you that because it took a long time and that's also the virtual machine that I'm using I will say that downloaded the Cintas a I saw DVD from this link and I used that because I had to GUI for the presentation and if you just follow the god yeah yeah if you just follow the god it'll it'll tell you how to create a VM and we're getting ready to create a whole bunch in a little bit yeah so a couple questions in chat already is just using the helper node or is it self-made you're not using our helper node or questions helper node right now I know that yeah I saw the two extreme when they help her now that looks pretty slick I'm not using the helper node and I'm kind of did that on purpose that way that anybody could come in and adapt this to their environment without having to have it scripted right that way it's kind of you know you're not too worried about proxmox or liver it's kind of all Universal escape it's kind of the plan yes so this is you know compared to the content that you have had on the stream this is definitely very basic yeah so next question would be why to use pfSense because it's got a GUI it's popular with home lappers oh yes I've had that question extends to me now you could definitely use the okt for services vm for DHCP and as a router but for folks you're just kind of coming into it that's a pretty complicated setup to start out of the gate so I figured this was a lower barrier to entry is really why I did it this way nice not greatly appreciate that so I'm just gonna run through a regular standard CentOS install I documented using the standard partition you don't have to do it that way but I like using the full storage space basically it just removes your home partition and then you use all of the space for your root partition so that's just documented in there on how I did it and then when you're running through that install you select the server with the Dewey and install the guest agents kool hain and then there's the two network interfaces I have one of them is on the my home network and the other one is on the ok D network and then begin install and then you just set your root password and create yourself a user with administrative privileges and because that takes so long I did that in advance appreciate that yeah and then so just this is something I do as soon as I get a new VM updated I installed Apple because I will need it later for xrdp and then I rebooted the Machine and then specific to my machine I installed xrdp and tiger VNC and these are the commands I used to do that and enable the service and add the firewall rules and that was just so you could log in and show us that's right yeah that way in zoomin yeah that way in zoom I'm just sharing my RDP screen so you can't see all my proprietary ECU hidden stuff [Laughter] so I also downloaded Google Google Chrome and installed get gets necessary Google Chrome for whatever reason in 6.5 and ESXi 6.5 and the web GUI there's some sort of a glitch where you can't see like the last virtual machine in the list so I installed Chrome and that's where we are at the moment that's the current state nice nothing else has been created so we want to go through and create this kd4 pfsense VM so you can download the ISO from here and then upload it to your ESX host now I chose you know the ISO in the mirror and this is we're just gonna go through and create the virtual machine so I'm gonna go to virtual machines and create a new VM and for a lot of folks this is probably gonna be a slow part something they are now to do but not everyone knows how to do that so we're just gonna go through it so I'm gonna select other and then Frias FreeBSD 64-bit we just use that data store and I'm gonna use all the defaults here with the exception of I'm going to add an additional network adapter something something to note is that this is network adapter warrant for okd we actually want the vmware network to be one because it's gonna be our way in interface but for some reason when you hit next here it will swap them anyway so it just it works out and then so now I'm going to the data store where I have the ISO and want to select the PF since ISO and want to run through the install and we will start that machine up and run through the install so well that's booting let's just go and look and see where we are with the guide so we've done that what created the network adapter and then we want to run through the pfSense install well that starting are there any other questions you see in the chat so a tree Piper's he's basically asking have you run this lab with this setup and the answer is yes you've you've actually run this a few times you've updated it from four four two four five just for the stream and thank you so much the the not quite sure what a tree Piper's like question is like when he says lab what does he mean right like they do does he mean like something bigger than your home set up I don't know okay D versus OCP so we're doing okay I'm trying to reread the question have you done yes he for lab with this setup no not what there's no you wouldn't use OCP for for this you would use okay D for for this yeah basically what we're doing is a bare-metal installing virtual machines and then that way you can use your own environment so you can do a bare-metal install on proxmox or virtual box or hyper-v or libvirt or whatever so this is starting up okay so he was asking about ICP and then right yeah yeah sorry I was super confused yeah my bed that's totally my fault and now there's the delay in chat and they'll just add to the confusion yeah so what we just is ran through the pfsense install and I just selected all the defaults suite and this is you want to start in just a second and we'll go through the setup wizard GUI pfsense provides what again the the DHCP and in the router and the router okay got the traffic yeah so it's up and I'm going to enable this second interface is connected to that Network and then I'm going to pull up the web GUI and we'll run through that wizard map so the default username is admin and the default password is pfSense and we're going to call the hostname okay D for Pia sense and the local alright spell check is good yeah set the timezone and on this wayne interface page we're just going to uncheck this block RFC 1918 private networks defense and this yeah we'll use the standard laner dress and then I'll put in a top secret password and hit next and then we'll reload that so that part is complete so let's go to the back to that guide and see where we are so we just did these exact same steps of you know logging in and run through the wizard so I'll just scroll past this and next up we need to create some virtual machines we want to create the bootstrap the master the worker nodes at the beginning of this guide it's got listed on the specs that you need to use so we'll just this is going to be a bore and part a suspect of the are we gonna get the demo yeah we roll through here so Andy for bootstrap Linux and core OS and each one of these is gonna be the same to four virtual CPUs 16 gigs of RAM and 128 Drive that's on the right network interface I have been provisioned these trials before and not had an issue but I'm not going to do that today all right so that virtual machine got created what we're going to do is just start this bootstrap machine and then when it's the console starts we're gonna hit tab and leave it at this boot where we're gonna put in the arguments and the deal is because we're not using vSphere it won't generate the MAC addresses unless you start the VM and we're going to need those MAC addresses shortly so let's create the to make reservations so let's create the okd for control one Linux OS so I had plant ándale and maybe doing this like priests it and a lot of this pre-baking it like we had talked about no didn't feel this genuine and it didn't take as much time to do it so I just kind of well I mean you know we've got now until no I don't have anything until 4:00 p.m. Eastern so yeah my time zone 20 time so we're sure he is or will he the same that there's a core OSI ISO maker their core OS maker ansible playbooks I'm not seen these before will lead he's saying that those helped him set up to a lot faster you you talked to Willie before right Craig yeah I've seen him in some of the okd working group meetings and I've seen a little not slack so now we're gonna make the second control plane how Christmas contribute to this that's cool sure it is yeah Chris mirandos I am I fell over at the hatter who's on pto this week nice and we will select that same Fedora core OS I say and if anybody in chat sees a mistake please feel free to shout out [Laughter] we're having important to point out that this this show would not have happened without OpenShift Commons I went to OpenShift Commons meeting her at okd meeting and Craig and I met through there and you know we've been talking on and off ever since trying to get this scheduled and we're finally here so yeah you shout out to the openshift comms community and I am for all the hard work she puts in from there yeah I guess I've been in there since and I hadn't seen you know I had so here's the thing right like every time I try to go to that meeting it's like something's popped up right before that that's like you know oh I go to this now so happens we're going to do the compute one node and then after this we only have one more left and we'll be able to do some stuff nice so you know you're you're working at a ECU you have clusters about I'm sure you know what kind of what what spurred you to build this home labs set up I guess this is a very good question and I will answer that in like five minutes because those are low that I have plenty of time to talk about that perfect so yeah let's see chorus ISO maker sounds really cool so while leek Christian introduced will lead to choreo iOS maker so I assume maker sorry I can't speak today so yeah if that's a shortcut around some of the stuff that folks in chat want to try highly encourage it you know I mean write a blog post come on the show all right so now we've got all of our beams and they should have all MAC addresses oh yeah so I didn't show this so to download the the ISO you can just click on the link and I'm a fan of stable builds I like those and if you go to bare metal and virtualize that this is the link to the ISO so I just downloaded that and then uploaded it to my host data store and that's the I saw that I was selecting for the the CD drive so this is me going through this is exactly what I did and this is what we've ended up with so now all we get to set up the DHCP reservations which is another boring part of the demo here so we're going to go into the pfSense VM and create the reservation so to do that we're going to go to the services and DHCP server this one right for me because I've never seen pfSense before so it's pretty hot yeah so I'm gonna change the range from 10 to 99 because we're gonna use the 200 and I don't want our servers in there and I'm gonna go ahead and add this DNS server and this is the box that we run when it doesn't currently have that address but it will shortly go ahead and add that I'll say the changes now I'm gonna start adding some reservations so I'll start with the bootstrap node and to get to the MAC address you can click oh great on so resize my screen it's a little bit different so you can click on the plus beside the network adapter and we're just going to copy this this MAC address and then we'll scroll to the bottom of this page and I hope I'm not going too fast but you'll be able to replay this video it'll be right here for 60 days after it goes final and that's always on YouTube ok that's alright so there's the MAC address the IP address and that's gonna be the host name and we're gonna say that cool and we were gonna do that for every other VM which is awesome I mean the pfsense though it looks like it's little - let's check that out oh yeah it's also talking about a tool and check called gov see I mean all the stuffs new to me what is go see yeah that's I think it's like a command-line tool for VMware where you can I think it connects to the API and you can create the MC so you could definitely use that oh yeah terraform there's a ton of things to make this a whole lot faster than what I'm doing it but yeah I mean you can use a playbook whatever you want yeah all right so that's gonna be two of one that's gonna be the control plane so yeah this is a little painful going through this but I think it's necessary pain is necessary sometimes ya know education for exactly so let's get the MAC address on this one copy to the clipboard this is gonna be 202 control plane t there's nothing really that much down yeah there's some argh from crisps and the ECU itcs family open chat oh okay appreciate ya watch no face 21 is the username if you know that one can you bein that guy actually I have a little I have a little Easter Egg up here form actually if you look at my virtual machines I got this pool house wife IBM right here oh that's the inside joke specifically for that guy so he's well prepared beautiful control plane three it's how we do have a good time a see you know I'm sure you do PC is a fun campus it's a good place to be alright see we got that's control plane 3d he says nicely done Craig well played sir so copy this MAC address for compute one and if anybody in the chess all errors in those MAC addresses now would be the time to tell me no one said anything yet all right so we're gonna make this YouTube one and we were almost done I appreciate you hanging in there with me oh this is this is nothing man this is easy this is all part of the fun gotta get to the good stuff somehow right this is gonna be all right so now that that's all done I'll hit apply changes and we'll go back to the guide and see we're just kind of scroll so we've done all this and this is what we should end up with as our MAC addresses fully populated and now we need to also configure last but not least the okay D for services via MAC address specifically the one that's on the okay D interface so let's take a look at that VM and this will be the last one so it's got two interfaces and the one that's on the hook ad network we want to assign reservation as well nice and that's going to be 210 oh we're getting close to getting some action here I'll save this and apply changes all right let's go see where we are in the guide so we've done this there's the O key D for services so we want to open terminal mo clone this git repo okay and just to take a quick look at that Gary [Music] what we're doing is pulling down some files these are pre-populated there's a H a proxy config and then it's already set up with the all the nodes and the IP addresses that we've used so if you're following through you're using this exact guide and the IP address is you don't have to change a thing we've got some DNS settings for our bonds forever we're getting ready to install and again it's already pre-populated nice and nice full repo yeah yeah I've had a couple at least one person commit to it your fingers dollar now yeah exactly right now installing bond and bond details for our DNS and then I'm going to copy these files to the right place that we that are in the git repo that I just downloaded and we're going to start enable and take a look at our bond status to make sure our name deesser status to make sure everything's good so everything's active it's running we are in good shape loaded Wow ma'am oh yeah like you it's almost like you've practiced this so many times all right so I put in the firewall rules and reloaded the firewall awesome one thing I need to do when my being here is I need to change the DNS of that interface that's connected to a home network I'm and that burned me once or twice so now back to the guide we're gonna restart network services the network manager and then we're going to do a dig we want to dig ok DDOT local and that looks good and we're gonna do a dash X on that - team alright cool so our dns is working it shows the api's we're gonna use so we're load balance in this on the same server so everything was good there then back to the guy this is a picture exactly what you just saw so we're going to install a che proxy I hope everybody is not bored now this is great so Jacko from over on the YouTube side of things says from his experience you don't need DHCP you can just run f costs ISO and then configure network by network manager next uses so-called built-in core OS installer with parameter copy network I'm not seeing that flag before so that's interesting I've to dig into that so I just copied the H a proxy configured it came down in the repo to the right location and then we're enabling the service we're starting the service and we're going to look at the status and everything is active in green so we're good to go on HTTP proxy except I forgot to add the firewall rules so let's do that real quick now that's important and Apache let's go ahead and install that best proxy for docker is traffic maybe that's that's pretty slick I mean traffic traffic or however you pronounce it is very slick yeah yeah so many so many choices the I liked Rafiq I liked caddy now that it's gone open there's also an tried-and-true of engine acts so you know I mean there's there's a bevy of services out there notice you use H a proxy and openshift yeah and a lot of the rail stuff so I would just keep it with H a proxy yeah we'd like H a proxy a lot we put a lot of work into it one one thing I did I didn't mention is I changed the listening port on the HTTP kampf to 8080 from 80 and I used I you said you know if you use nano or VM you're more than welcome to do that so I've started the service and put in the firewall rules and we're going to a curl and make sure that the web server is working and we got a test page that looks good we said we'll keep keep rolling all right congratulations you're halfway there download the openshift installer client so if you go to the github open shift okay D releases page the latest release is a 4.5 beta 6 that Vadim cut a few days ago yeah and what we're going to do is download this openshift client and pre-shift install client X cubed for Linux so I've already got that in here so you can just copy and paste it and I try to keep up with that with the latest versions yeah I always wish that we had like some links to latest and everything right like : latest and the the container world or hey like it we just had like a sim link to the latest version of you know whatever installer we wanted to be cool but yeah all right so I'm gonna extract those files that I just downloaded and then I'm going to move them to the user local bin path and then just double-check the versions to see that they're good so that's good so they're both using the beta six version of 4.5 the same dealer so everything is copasetic there if you don't have a ssh-keygen now is the time to generate one and I do not try one we're just gonna go roll with the defaults for this and we want to edit this config dot animal this is another file that we that came down in the repo and if you have a Red Hat pull secret now is the time to use it I do not or I'm not in this guy gonna use one you put the Red Hat pull secret here I'm just going to use what we got but let me catch my SSH key copy that and if folks aren't familiar you can get your pull secret from overshift comm slash try your SSH key though as you saw you have to generate that yourself yep so we're gonna paste that in here huh if you're unfamiliar with them hi goes into insert mode excape when you're done : name Debbie Kieffer I quit already the most popular video in the world now cuz you just showed people how to quit them it's uh I mean it can be complicated well you know what what's funny is as uh you know I usually open them for my son and let him just smash way to keyboard sometimes but the first time I ever did that he got out of them and was in the terminal so I was very impressed that's awesome [Laughter] alright so I'm making a backup of this this Jamo file always back up that install config file always always always I'm not the only person who does it that's good yeah create some manifest and what we're doing is we're using this open stall openshift install file and we're pointing to the install directory that has that install configure MOU that we just edited and we run that then after that runs we want to modify this one file this cluster scheduler 0-2 we want to change the masters schedule true to false so that pods aren't being scheduled on our control planes so I'm just going to set it you can vim it or nano it if you like and I'm just going to Pat it and make sure that it's yep so master schedule is now false all right so now we're going to create the ignition configs when this is done look in that directory and see what's happened alright so we look in the install directory what we just did is we generated these ignition files the bootstrap was gonna go to the bootstrap node the masters are gonna be used for masters and the worker is going to be used for the compute nodes and then there's my back file it also generated a list an install directory that has our cube config and our Kiba admin password that we will use at a later date yep so okay if you reuse if you try to rerun the configs for some reason if something goes wrong in your environment and you don't delete the directory delete the install directory there's some hidden files in there and it's caused some pain in the past for me so just delete it and start from scratch all right yeah all right we're going to host those files on the web server that we're on so I'm going to go ahead and make an okay the for directory in our web root and then I'm going to copy those files over there and set permissions and we're just going to test the web server and make sure that we were able to curl that file so that works we're good also you need to download the core OS bare metal file so they and that's on the same page you go to bare metal and you can get the raw file here we're gonna be using that and also the signature file you'll need that as well so I've got all that copied here and what we're doing is downloading it downloading both of those files we want to rename them to shorten them up and then set of permissions and this will take at least 70 seconds or so okay on my high-speed ish Internet so Beautif you got questions well there's there's a lot of activity in chat I'm trying to keep up with people talking about Venn vs code I don't think there's any questions necessarily but we can pause for a second while this is going and you know what the chat catch up to our voices here but yeah let's see just exiting them is way easier than it used to be nowadays you can search for the Stack Overflow close that explains how to quit that it just killed me when I saw that he said that was antimatter yeah that that that is awesome so yeah that's good stuff if you so if you install ok t4 in a closed Network remember to add the proxy and no proxy settings in your install config file that came from Jacko on YouTube I'm not sure Langton and then what about in them so that's down alright so now we get this is another boring part there's a lot of boring parts of this but we're going to start to do strat node and we're going to type in these these arguments mainly what we're doing is we're typing in the installed device which is going to be SDA and we're going to point it to our install image which is the F cos da role that we just downloaded mmm and then we're going to tell it to use the bootstrap ignition file on the bootstrap node and then the masters are gonna use the masters a network responding sword right so let us pull off the bootstrap node and in another window I want to pull this up and this is where typos will really hurt you so I'm gonna pull it up so I can see make sure I've been the right thing here alright so I'm in the bootstrap node and I'm gonna do chorus I just assumed the dynasty means install but I wonder if it doesn't mean instantiate I thought about that earlier today huh maybe I don't know yes with you one of the minimum memory requirements for this lab for Oh kitty for you said you were using 96 gigs of RAM that's what I've got yeah if you pull up that spreadsheet you can do the math I think it's like 60 like 64 I think you need 64 like that's the bare minimum but I would not just wrong with the bare minimum to be honest with you folks the more the merrier right starting to sound like a developer I know I know I know I've been hanging out with developers too much all right so we're putting in the path of the bootstrap in the file you just do a quick check to make sure that's right all right so we start the bootstrap node I usually wait until it gets to the point where it's download it confirms it shows it's downloading those files and then I'll go to the control plane okay is this the point where you wanted to talk about why you did this [Laughter] we're getting close to that I appreciate you reminding me of that nah no problem okay d4 B when will okay t4 be in a gea state I believe like tomorrow it's like soon I think I mean I would say by the end of the month yeah like soon I know that I'm not saying bagging a little bit oh yeah I mean I always try to be a little you know not nefarious that was the wrong word but a little ambiguous about dates because things do happen security dates and you know we only release software certain days you read at various reasons for that - and so a lot of stuff goes into a release for something like that so yeah our next version of opt for is coming soon I mean I could get you a hard date if you really wanted it but it probably you know wrong so unless it is indeed tomorrow but yeah I don't know when it is supposed to be GA officially to be honest with you folks and the soon is the best I can give you sorry so I'm just scrolling down this guide and so we did the bootstrap mode and then the control play nodes and just you know changes in the this ignition followed a master on those so that control plane one started at the one do control plan to couldn't the start parameters also be copied in the Machine configure the VMware and yes that should be yes I think I've seen that in the OCP documentation yeah at that end that keeps you converted base64 I think there's I think it's a little bit more complicated than Justin yeah it's not you have to do a few extra things but it does prevent you from having to pull out this massive line yeah all those years of doing latest speaking teaches type in Spain off right now that's good so it would be interesting to watch okay D OpenStack with auto-scaling and Red Hat storage integration Amy if you are watching let's see is she spots on she might be but she might be in a meeting to knows but I will write that down that's actually kind of fun because we are trying to do oleg we are trying to do more open sticky stuff you're in the future there's been some requests for OpenStack things so I haven't looked at the stack person on my team and I let her know immediately is the free VMware server support multiple hosts added to it no idea no it doesn't you can have multiple versions of those but you can't manage them in vSphere got it these fear is not free no Narenda says I'm out of my league here I don't even understand what he's talking about so Miranda don't fear the reaper go to ok d dot io and you can start learning about things there but he's in vmware this is ESXi it's a typical virtualization host he's setting up individual vm's like you would in like a VirtualBox kind of scenario and he's installing the appropriate distro or ISOs on to the right boxes for ok D for antimatter says you do get a 60-day trial though with the vSphere and V Center with all the features enabled that's good to know Jack o from YouTube my question is can we use install config dot the mo the pull secret from the Red Hat OpenShift cluster manager site the answer is yes any other questions we need a tutorial on ESX that well mmm-hmm TBD we might have some fun with VMware here on the channel soon yeah you got stay tuned around here folks always got good stuff we're on our last two and then there will be a nice lull where you can ask plenty of questions and I can tell you my heartwarming story about the KD okay overt and rev as a platform please yeah so we need to get more rough stuff in there we'll each says if you are in a V mug you can get most of the VMware licenses for $200 for a one-year evaluation did not know that I'm not Ana vemma good never have been so so this is gonna be the worker cognition file let me scroll down and I'll just hit enter and then I'll scroll down in the god so we've done the control planes we did the bootstrap and now we are on the compute nodes last and we will be done with a ton of typing at that point nice and I can breathe easier huddles J put the link to the membership advantage thing in there for V mug folks so thank you for putting that in chat can we get Reb 3d versions or the Red Hat OpenStack free versions yes there's upstream versions of both of those there should be Rev is based on overt OpenStack is based on our do so yeah those are your two up streams for that [Music] there's a lot of chat activity I'm just trying to read through so yeah you want to tell us your heartwarming story yes yes I will I mean like I don't know maybe one or twenty more seconds and I will sorry to keep delaying you there oh you're fine I'm just about at a point where I can we're just gonna be waiting on the machines all right so almost there I won't go back to this guy so we're down to compute nodes or the control planes so it could be nights one thing I did want to mention this has come up in some of the gods is that the compute nodes will show this internal server error it'll show that until the boot mode the bootstrap process is complete so don't freak out just be patient yeah patience is key when it comes to anything kubernetes related right like things just sometimes just need the weight you know so they come up so we're gonna monitor this bootstrap process all right so the kubernetes api already came up so we won't wait for this bootstrapping to complete that might take 15 minutes or less I'm not sure how long that's going so now we got plenty of time to chitchat wait there are some other things we can do in the meantime but so to answer your question what was the question how did I get started yeah like what what made you decide to build a home lab like what took you down this road of okd at home kind of thing so I guess I guess Red Hat came on site and did an engagement with us for our okd for our hooking shift clusters and at the time I wasn't heavily involved but I needed to rebuild those clusters so what I ended up doing this was ok t-33 dot before ish I think version so what I ended up doing is got some equipment and used my home lab and I built the okay 3o Kitty 3.11 clusters and perfected my Hansel my inventory file in the Play Books and I will say that uh ok d4 is - way faster to install but so I kind of cut my teeth on the 3.11 in my home lab on purpose got that perfected and then I was able to use what I learned there in he's use production environment by already perfecting this play books and having to run through it and we had a couple unrelated issues and I was able to use what I had learned already it kind of just elevated my skill set a little bit by being able to run through it and tinker with it a lot so I'm a big fan of learning by experience and if you got the ability to to purchase some relatively it's relative to your situation hardware and set it up and learn that I'm all about it and that's really the whole reason I wrote this guide so I was a little you know when hook OCP 4 came out there wasn't really an upstream version available yet so we kind of had to wait for it which is which is fine it's not done that so I did oddly enough I ended up having to use the OCP documentation as a base for my god for the okd documentation and in the past it's always been the opposite but now that okd is the most GA I think we're all set so that's kind of how the guide came about and how I used it to answer your question if that answered it it does yeah that makes total sense some folks here are saying hey they want to see like the install running so there's a few commands I think you can run to just you know you can tail a log you can yes SSH into oil alright core here's another thing I'll show you if you SSH as the core user you know I just make a new journal there you go SSH is the core user to our bootstrap you blow up that text alone you say thank you perfect and then we can look at this boot cube service so we'll just let that run and I'm you know what I think what they were asking for we can put this in the log level info I think we can put this on debug mm-hmm I'm not sure if that's right I don't know if that would do anything if it's waiting let me up chat let me know what you want me to do look works according to rockhound all right so this is still it's still booting you know a low-tech way of checking to make sure everything's working now if you click on the on the bootstrap I'll look and take a look at go of course announce down here yep all right so it's doing stuff that's good the control planes Oleg Popov says I feel uncomfortable not seeing the logs that's doing stuff you don't believe it yeah all right yeah it says these compute nodes show this internal server error and that'll go away once the bootstrap process completes right so like it's a it's an onion right like if you want to explain how the the layers of the onion kind of form together feel free like you have to have the the bootstrap then the worker or the masters and the workers and then CSRs happen all kinds of fun stuff yeah all right so we're monitoring the bootstrap process and then once that's done you can comment out the the bootstrap node and H a proxy and reload the services and we will do that shortly while we were waiting if you're alright with it we'll skip ahead in the god a little bit and do some things that we can do any descent we need to set up we need installing fists and set it up on this okay D for services VM we're going to use that as a persistence towards for our registry and we're going to create a wordpress container and push it to the registry and make sure everything's working so I'm going to do that real quick let me create a new tab oh gosh all right so why not a kinesha NFS you know kanesha NFS is oh that's net using I'm not gonna say because I'll sound stupid if I'm wrong so I think it's using uh-oh I'm looking it up I've drawn blank use mustard maybe cluster so what I did is installed NFS you tools which was already there enabled the NFS server service and RPC bond and started them and created a directory and set permissions and then now I'm going to create an NFS export you can limit our Nano I'm going to do an echo to a tee to create the file and then I'm just gonna at that file to make sure so we just created an NFS export nice and we're going to restart the NFS server and put in our firewall rules and at that point our NFS should be set up and we'll create a persistent volume at once our cluster gets up and running so we can store the image there our image registry so yeah apparently Ganesha NFS NFS Ganesha we're the right way of saying that is it supports things like cluster and Saif as well oh okay that's where I've seen it alright alright so here's the wall what are the questions we got all right none so questions fire away folks so like what is your install base AC you look like can you talk about that at all where it's a very good question I mean here you're using OCP right like that's right yeah we've got a couple clusters one more than one more than one that more than one this is and we primarily right now use it to host WordPress sites for our departmental sites okay then there's a few other applications we ever running in there along with Jenkins as we kind of used as a utility so that's kind of how we use OCP right now okay world we're on 3.11 I got a request in to go to four so nice I'll be happy when we get that rolling how are you thinking about managing that process of going to Florida yeah I go under four okay I mean I got a migration tool right for little migrate that Matt Woodson was telling me about so I need to take a look at that but yeah we'll just have to create a parallel cluster a parallel publishers and then migrate those projects over yeah the the way we generally tell people is that uh build your pipelines to deploy over to those clusters oh okay nice hey I learned something today there you go but I mean that's you know that's the simplistic way right like it's not always that easy you know it depends so if you she set up your four clusters difficult over there and you know kind of just move all the infrastructure over there after the fact or build the pipeline and you can actually just run it in no kitty for OCP for there as well you know there's a lot of different ways to skin that cat for like definitely sorry for the cat fans out there so watching a face 21 says craig is being modest but he was the person behind migrating ECU's OpenShift from 3.4 to 311 and also moving from OpenStack to VMware with oak 84 what is his most anticipated feature now working with oak 84 I guess that was a question what do you like what is your most anticipated feature working with oak 84 think he knows the answer to this it's it's a lot easier to administrate I think this is probably the main the main thing so yeah with you know going from 3 to 4 we made a big investment and coral s operators you know we acquired coral last few years ago put a you know they created the operator pattern and we have essentially tried to operator operator lives operationalize the entire openshift install so everything is an operator so you have everything's being installed by an operator right now and okd operators are spending up different controllers and different things for the entire stack or all the stacks that run in under the hood of OC PE / ok d the operator pattern essentially it takes operational life you know things that you would do in those day to kind of know patterns of install upgrade manage those things get operational IR codified and then instantiated as controllers on the clusters themselves that way when it's a text an event it then runs a corresponding event of course it then runs and then take some action to make a declared state possible let's see I was listening to a podcast I think it was a cube CTL pot podcast it was I don't know a couple years ago I'm like I can't claim it somebody mentioned that the term immutable lifestyle yeah that would be a pretty cool t-shirt ya know it's I mean the idea is you put an operator in place and it just handles stuff for you right like you don't if it's if it's designed to manage itself like you we have a maturity chart basically a little 1 through 5 level one is just install level 5 is like fully automated you know self-manage installs and upgrades self manages everything if you've got you know a level 5 operators running and you're kind of just you know it is kind of like just keeping the lights on at that point right and you're just if the more operators you build the more you know infrastructure the more IAC that infrastructure code pattern right like operators really fall into that very well let's see lots of lots of chatter amongst people which is great like I love it I'm trying to filter it through the questions and they're enjoying the oh yeah they're always worried it was gonna be a little too boring no like see here's the cool thing about streaming like the way we do here so we use a service that aggregates all the chat between twitch and YouTube and Facebook so people can talk to each other across platforms it's it's it's really what I think kind of makes us shine is that we maintain a sense of community amongst all the different platforms yeah that's the symbol cast is pretty cool so so the bootstrap process did complete so yay for that so let's want to exit out of here and then we'll go back to the guide and see exactly where we are we've got maybe too many windows here going lead I have an answer your question I actually just got it not too long ago found out about it yesterday let me find it real quick lead asked wherever where to find open shift common slides so now that the bootstrap process is complete we're going to go ahead and comment out that bootstrap married see in the h8 proxy config okay just to kind of take it out of the loop if I can type in my password correctly and if you look here you can see where it's commented it out I you said but you can be MIT or nano it if you like and also while I'm at it I'll just go ahead and shut that virtual machine down just powering off it is no longer needed cool so yeah is there a need for additional external cluster monitoring solutions such as New Relic or cystic with O's TP for um to ask that watch no phase 21 so let you answer that yeah yeah so I mean it really depends right like you can it's you know we have operators and OCP for to set up an entirely you know new F stack which is elasticsearch fluency and kabah and that'll give you all the graphing and everything you want for your applications right there that's an operator from operator how that you can install and then use it that way you're using the same tooling that's you know baked into the cluster already with you know your existing monitoring stack but it really depends on your preference right like so New Relic insist dig that's gonna be you know well syste giving that's deep applicate like that that looks at crap brain dumped the whatever service it uses in the kernel to look at things like sis calls and everything yeah so cystic is gonna give you a very deep dive and it what's happening in the cluster but whether or not you need that it is up to you as far as like sorry sorry about that as far as uh what was the other one New Relic I mean New Relic can do different stuff like you know applications monitoring and you know uptime and all that fun stuff so it depends which components and you were like you're talking about if you want to bake New Relic into your App Store whatever for APM I think that makes sense right like there's there's also a lot of other tools for that so yeah like we we partner with cystic there's a statistic operator ebps thank you a lead EVP F extended Berkeley packet filter V that is extremely powerful and that's what cystic utilizes to look at things under the hood the the there will be a stream in the future with someone with cystic I think we have one in the past O'Keefe Commons with cystic as well so stay tuned for that alright I'll say that so that the bootstrap process completed and so what I did was exported the cube config to this environment variable and then I was able to do an OC am huh so I'm in as the system admin and you can do notice to get notes you can see that the control plane nodes are up and we also do see get yes are for certificate signing request and we're going to look for some pending certificate signing request and we want to approve those those are going to be coming from the compute nodes so if you look at the compute node now it doesn't say the you know it doesn't have the API error because it is work which is positive one other thing and I guess maybe I'll put this in the tutorial backwards but you can use JQ very like a ton of certificates at one time so I'm just going to copy and paste that in here and what we're doing is we're downloading JQ we're making it excu and then we're gonna move it to our user local bin and then we're gonna test the version so it is now installed and I'm just going to look out before the certificate signing request because he gets and there's two pending it's kind of hard to read so what we're going to do is approve those with this JQ all right let's do this so that's going to approve roll okay I see get nose and there's no chart there yet the workers but they should be here shortly because if we do another hosi get CSR now there's two pending one for compute one and there should be another for compute two here shortly there guys I compete too so we're just going to rerun that same command and a producer tick is there now if we do those he get nodes there's our compute nodes so they're still starting up yep this is a good time to do an OSI get cluster operators yeah and see what we got going on here so these are man I wish I could you can shrink down your font if you want to I mean I know there's it's harder to read for folks but getting it all on one there there's a happy medium between being productive and being able to display some time so right so these are all the different cluster operators and they're in the process of starting up so if you look under this column it'll say available processing degraded and it tell you how long so a lot of these are coming up what we need to do I need this image registry to come up and we can skip ahead to that a little bit and then after that once the console comes up we will be in a good spot to continue with the demo but for right now we need to ya the console will be one of the last things to come up yeah yeah that's classic yeah I mean it arrives on top of everything else so let's see where we were with the guy so we've done that we can we've done the OSI get ooh there's a typo let me make a note on that we see a lot of people a lot of people are thanking you for this guide in the chat by the way I really do this is so awesome it is so awesome I'm happy I tell you it like that's why I wanted you on because I knew people would eat this up like well when I tried to build you know I was I heard about I tell what the number one thing about okay the or even OCP for was was to over there updates right the you know if you use core OS is your base platform then it kubernetes can manage the cluster and the nodes and and they can do its thing so the amount of hands-off is very attractive to me because it can be kind of taxing yeah so what that's that's what I wanted and I forgot exactly where I was going with this but that's that's how I got into where I wanted okay D for to be right the the over there updates is the big thing about OCP for that I think a lot of people kind of forget about right like I I if I spin up a cluster right now for for cluster and our you know dev environment that we use for you know beating up demos and such it'll give me like four four four four five I think is what it gives me and for for ten is out already and I just click like the blue arrow update and like it all off it goes and it just starts upgrading pods and operators as it goes so it's it's really nice just to have that push button ability to just upgrade and get the latest stuff and it'll just let it as long as your applications are designed to have multiple pods and can be you know managed in the cloud native way your entire upgrade can be non impacting which is pretty cool words reals yeah we're trying to get to that point where it's just like folks are just in the constant mode of updating their clusters right I'll say that what happened para what happened was when I was trying to create this guide you know see only the documentation for OCP was available and the hardest part about getting it working was creating the environment the entire environment and then so what you've got to have is you've got the DNS is the tricky part you've got to have a load balancer there's a lot of pieces that folks who are just starting out or that run home lamps may have not gotten to that level yet yeah which is kind of why I wrote this the way it is and it's kind of very simplistic compared to the content that you guys put out on this channel but hopefully it'll help some folks get involved and and get that a local shift ball or the okd ball rolling where people are getting out off with it a little bit easier I definitely think some people have learned a lot on this on the stream so far so thank you so much for joining us really appreciate it all right checking out this cluster operators we are still go and waiting on that image registry what we're going to do is when that image registry operator gets up we're going to ex I guess we can go ahead and create the persistent volume so let's scroll down to that so what we're going to do is create a PV if you do Oh see if PV we don't have any persistent volumes out there not now and what we want to do is create one using this template file so I guess before I do that I can cut this template file and this is just a regular template and persistent volume I'm gonna name it the registry PV I've got it set for a hundred gigs of storage we don't have a hundred gigs we're using FS it doesn't really it'll just use up but it can use other you got now if you're using block storage it's a little bit different we got readwrite mini and it's going to retain policy and then this is where you specify your NFS path and hang your server so we do a hosi create dash F and then point to that file and then do IOC get PV our Peavey's made its s status painting it should say available shortly so once that is available in the image registry is up we'll edit that operator and can point it to that PV and then our images can be stored in said registry beautiful let's check out the cluster operators and see where we are getting closer but surely I mean but the thing is is that the all these operators right like I can tell you what each one does but the a lot of them are self-explanatory too but the the the open shifts of lifecycle manager that gives you operator hub right like that's what makes a lot of V I want to install a monitoring stack on my cluster oh that's just a push button thing I want to install a monitoring stack in just this namespace oh it's a push button thing that's what operator have gives you and OLM operator of lifecycle manager is the piece that keeps operator have up to date in the background or that I should say the the package server and catalog piece but the OLM piece is the overarching project missed the beginning what's the basic setup here bare metal Cavey invert it is ESXi on bare metal yeah well scroll it in so a little picture here oh thank you that's great so the the setup is were using ESX these are the different beams that are created the type in the OS you know the resources and IP addresses and there's a look at the network layout that's pretty much the setup we're basically doing a bare metal install on virtual machines yeah essentially I hate to repeat myself but just the reason that's done is so that it's kind of portable you can use it on your own environment you can use the same method on hyper-v or proxmox or for total box for liver Greg med made this as flexible and simplistic as possible intentionally for home users that might not be fully versed and like a tariff or an ansible you know some other automation framework yeah there's also two sides of that story I mean I'm also not extremely terribly versed myself it's the same level with some other folks so that's also part of the yeah no no no I get that question are you using pixie with legacy BIOS or EFI the stream is hanging I'm not seeing a downstream yet see looks good to me yeah if it's stream singing for you just hit refresh y'all so the question was pixie with legacy BIOS I'm assuming that's what you're doing no no you're not doing a pixie that's right you didn't say yeah yeah there's pixie yeah that is just a regular buys there's an idea on these beams a check yeah straight-up buys question I've got to leave did you fix the authentication operator I'm not sure what that question is in regards to mark Marty sorry I think at one point in one of the earlier versions I'm not a thousand percent sure on this but I think there was the issue with the Authenticator operator or the authentication operator and I do believe it is fixed because here shortly we're gonna add a an HT Pastore provider operator so we can Logan is someone other than queue bad man let's take a look at these closer operators taking their time you can just throw a watch on that if you want all the nodes are up that's good and all these are approved and issued see and check out this guy and see where oh yeah the PS if the pv still says it's bending no wonder which operators are on you should say available so that is concerning storage isn't up yet though storage operators not online are you know are you waiting on that possibly yeah that's what I'm thinking let's see any more questions y'all a bunch of y'all hanging out here let's see so an OC get closer operator output there's unknown for authentication is it yes right now it once the cluster gets up and fully runner an operator will these I think I'll come explain that well yeah all these are still slowly coming online of course you know I don't know bring this I don't know how many times and of course this time it appears to be going the slowest okay core is there I just saw something yeah our image registry is now up cool we check that where's that storage operator better at the very bottom it's only been up for seven minutes I've been working for seven minutes I should say what are the big feature differences between Oh KD & Oates ap oh my gosh uh-huh well I mean okay D is the upstream for OCP willie the slides not sure there's no slides for this one but there is the blog post we just I'm asking Willy this will lead where you're looking for slides firm something from yesterday I'm still waiting on them so sorry people and actually I had a discussion about like just creating a github repo for just dumping slides in there or something in the future so yeah alright so the that image registry operator is up let me go back to grad the image registry operators up and our storage operators out there coming along slowly so earlier we did we created that PV and then we're going to edit the image registry operator to point to that PV so right now if we do a OC get PV which I just did it's available and now we are going to edit this image registry config and what we're going to do is change the management of state to removed and we're going to add the PB scenic claim this that came straight out of the OCP documentation cool tendency so using a state we're going to change that so sorry offer insert managed and then storage I'm get rid of these curly brackets gonna do two spaces and do PVC to a new line and then do two spaces after that Hindu claim C hit escape and then wait just a few seconds and then do a get PV and hopefully it will say claimed bound all right cool so cold now it's bound so what it was is that registry image registry operator was looking for a claim it was the only available one that met the criteria it was the readwrite access mode and was at least on her gigs or greater and so it bound to that oh so we were doing pretty good so our images will be stored locally on the NFS share oh a lot of these are coming up cool yeah we're close we're very close very very close yeah I think we're probably two minutes away from the console coming up and then we'll be on easy street yeah my question after this is from Jay echo again on Twitter after installation open shift sample operator is always degraded do you know what the reason is what the impact is for the whole day okay D so I think there's a I think it's there's a github issue 167 now in okay deregister where they're they're trying to get that cluster or that sample operator and add some fedora images to that so if you feel like contributing you can definitely go to that you said 167 I believe that time ago I was gonna do the same thing yeah come on come on Chrome you can do it so it's here there we go yeah 167 I'm so good padeen put in a an issue to request with that and after I'm done with this this demo this upcoming we complain to try to help out on this but I think that the operator at least old mine doesn't show is degraded but in the previous ones that did we will find out shortly yeah console all right so let's go back to that guide and see where we are because I kind of lost my spot we can actually go ahead and create this WordPress project without the GUI being up yes we can so what we're going to do is make a wordpress project so I'm gonna do an OC new project and want to call it WordPress test and then when you do that it will switch you to using that project so what we want to do is create a new app in that project and we're going to use the CentOS s2i image with PHP 73 and we're going to point it to this the WordPress get repo and what's going to happen is when that image builds it is going to take that github or that WordPress github repo code and put it in the Webroot and then build the image pushing to our registry and then we should have WordPress also I created a route which all that does is create like a router it's like I don't know word press dot ok d dot lab or whatever it's going to be and then while we're edit we're also going to add a Maria DB container cool and with some environment variables here and the environment variable are the databases gonna be named WordPress to users WordPress and the password is WordPress now this container is ephemeral both of these are I didn't setup persistent storage on these just to show you what it looks like but yeah this is a demo if you turn this off it'll go away all right cool so our console is up so let's wait that's perfect timing because I think the next step is open the console which I don't have a link to that's unfortunate all right console and then we need that cube admin password too so let me get that cube admin password that is going to be install files off so this should be a look at the web interface or gonna log again is the cube admin user paste that in and this will be our overview our cluster is still updating this will eventually turn green right but let's go ahead and look at that let's go to the developer window and look at the WordPress project WordPress test project we just created you know this topology view you'll see that the Maria DB containers already up and the WordPress container already built which was concerning lis to quick we'll see about that yes here's a look at the boot yeah that should have done that fast so anyway it uh it pulled down that image and it put in the wordpress code and then hopefully here somewhere we'll see where yet so pushed to the image registry so in just a second we'll see that that hopefully pushed to our NFS tour alright cool so that's running so we'll go down here to the route and there's - yeah so we'll just go ahead and figure this while we're at it yeah why not and then I'm gonna use the service name which is ma RI adb ok and then run the install and then we're just gonna call this all these I know that's your email address yeah yeah yeah install WordPress disses and Tans and then we're gonna login to WP admin and we have a wordpress installation and once you get to hello world you know everything is copasetic yeah so Wow as part of that we're going to look at we're going to add EU SH one for NFS share registry and now there's 272 Meg's that's all damages that are being pulled down for everything that's running in cluster right now well this is what this what this is is the image that we just created with WordPress it's just that just that oh yeah yeah so I've been there no worries in there see there is now a docker folder which the image registry operator created for us and I think I got one part left of this demo we will be doing a blanket ask why is it always WordPress because everybody uses WordPress like just a number one let's see so this is our we've gone through all this we did the WordPress thing we got the hello world and last but not least we're going to set up a this is a saying yes 200 now we're going to set up a identity provider so let's just go back here so the cube admin password kind of really I think meant to be a temporary yes very much so so what we're going to is use the HT password and we're going to create the test user with the passwords test password and we're putting it in this users HT password file and then we were going to create a secret in the OpenShift config this - and selects the project so we're saying create this project using this file and the OpenShift config project so we'll move and then forgot to hit the old copy button yeah I do that all the time all right and then we're going to apply that identity provider now this is a template that I set up that's in this ok T files folder it's got that and I think I just got this off of the CP documentation know wouldn't spread like it's probably in the kubernetes documentation who knows yeah there's probably a few places you can find that and then now we're gonna log out and this take this may take a second to populate so right so right now it doesn't show it okay but it will Sam I yeah the name of the game with uh at least this demo is patience yeah I mean it's not like you've got some super high speed data center running in your house there you don't have AWS running in your home come on no one does that's that's why we're doing this cuz setting up on my ABS is something we have to do nowadays Lincoln says he uses he uses WordPress in his demos too it just cracks them up but it's always WordPress that gets used yeah I don't I don't know why just is so all right so now that that worked the HTTP password provider so we're going to log in now is that test user with the password is test password and so now you're in is that user and you have access to nothing what we want to do is give yourself cluster admin rights so yes as we've seen uh projects and just want to copy this so what this does is this adds the cluster role which is cluster admin to the test user so we'll just paste this in and then immediately you should have all the projects so now you can log in not his cube add me and as your self and tinker around with your okay shift or your okd install yeah all right so that's that's pretty much the ended the demo this is awesome I mean this is great yeah I wanted to say that you know so we pointed out that github repo if you've got issues you can create them there there's a OpenShift users channel on the kubernetes slack you know a lot of us hang out there if you get into a problem let us know there's the ho kitty working group they make bi-weekly I've enjoyed that quite a bit a lot of the usual suspects that are there also in the Kate slack and then there's a Google group for the hook ad working group and that you can join if you want to get involved and that's 233 so we're hour and a half in that's all I got brother and that was great that was extremely fantastic is what I'll say that was awesome now that I've logged in as the test user and I made an administrator the overview oh yeah buttons go on see you accept the logout and my back in I don't know how big of a bug that is but that's that's a weird ux bug to figure that one out now when you log back in there it is I'm hoping look at there hey look it's completely green what is here let's cut it off now so nothing turns red yeah okay run the bumper go now this thing I mean it's gonna stay right like it'll just slowly just stabilizes and everything will be fine right like the activity slowed down here and if you're not doing anything on it yeah so at this point if you're running this God and going through his demo in your home lab you should have a decent working ok d5 cluster that you can tinker with and teach yourself on how to do other things use another blog post there's a the openshift blog blog Pacific amor yep it's pretty solid they've got some really good stuff oh yeah Alex handy does such a great job getting content out there every day for folks right like I just looked at it yesterday and psych oh my gosh feel like it feels like you know there's like a bajillion articles coming out every day there's so much activity in the cloud native space right like if you go look at the landscape for CN CF I mean it's we can make a joke about it we could say it's just huge but it's it's it just shows you how big the community is and how many options you have but yes there's a lot of thanks in the chat especially from JD Farlow will this upgrade via the console to four or five when released this is our already four or five so if you installed before four yeah you should be able to do open the console and do an upgrade well he might mean is it since we've got the beta 6 when the GA comes out this automatically upgrade oh yeah in theory you just flip to the GA build I don't know yeah exactly I'm not qualified to answer that I don't I don't know the answer to that question unfortunately I think I think they may cut a release candidate the word on the street is that they're mine I could or at least candidate within the next week within the next maybe nine days or something so I check out you know check that out when it happens if you want to be part of that github repo okay D okay okay D if you want to watch that you'll get notifications when it happens and it'll be a good time I'll leave this up and test it and let you know if it if they updates to the release candidate or not and I sure that works out for you I think you would have to like so we have the concept of like Knightley's and latest stable whatever you know and OCP i think you would change the update channel maybe hopefully but they would have to build in a caveat for that so I don't know yeah we'll see I would hope they'd make that easy but yeah this has been great so Craig we're mean we've we've posted your blog post a bazillion times but if if you've you've talked about where to find you is there anything else you want to mention on the stream I mean you've done an amazing job here today thank you so much for coming on yeah no problem yeah that's it if you need me I'm on LinkedIn and slack and that's about it all right well thank you so much Craig you know this has been a fantastic demo really appreciate everybody for the questions the chat it's all been great thank you so much for the follows everybody that's followed during the stream you know I am done streaming for today but tomorrow we have OpenShift commons coming up with a deep dive into spinnaker ooh the spinnaker operator that looks fun and for the day that might be it tomorrow Thursday's are usually light on streaming Thursdays and Fridays are usually lighter for us so yeah but tune in for openshift commons at noon eastern 1600 Utes 1600 UTC tomorrow and until then stay safe be well and enjoy watching the stream here in the future thank you so much [Music] you

Channel: OpenShift

Views: 12,323

Rating: 4.7333331 out of 5

Keywords:

Id: qh1zYW7BLxE

Length: 98min 24sec (5904 seconds)

Published: Wed Jul 08 2020