Integrate Myriad Default with airSlate SignNow

Save myriad default

[Music] hello everyone uh welcome to the session today i'm going to talk about how you can deploy spinnaker using home charts there's a lot of ways you can deploy spinnaker and for production deployment it is recommended you deploy spinnaker in a kubernetes environment and you know there are operators available some people just install helm and then you know go through the process of installing all the spinnaker components but i'll show you how you know you can use help to very quickly install spinnaker like you know you can get a basic spinnaker up and running in less than 10 minutes and then we'll go through some of the customizations that you can do with spinnaker i'm nermalya sen uh i represent a company called ops mx i've been in the industry for quite a while uh 25 plus years uh and uh you know quite interested in the cloud adoption growth and how you know deploy deployment to cloud environments can be made more seamless smooth and at a faster velocity before i go into the session let me spend 30 minutes to talk a little bit about opsamex who we are we are an experienced team that have worked in the past and companies like netflix cisco oracle aws we have 90 plus employees worldwide uh and you know we provide services like you know 24x7 support of spinnaker managed spinnaker and then uh you know onboarding and design services around spinnaker we are very active in the community in spinnaker kubernetes jenkins we have active contributions committers and you know we love the open source community with that let's get on the agenda uh what we're going to talk about uh today um again not too much of slides um like uh this is a live workshop so i'll really uh do a show and tell mainly not slides uh so we will walk through like i mentioned earlier um how to do a default installation of spinnaker using help uh very quickly and then um you know look at all the customizations that are provided uh within the framework so that you can do a lot of the customizations that are needed for a production deployment of spinnaker you know including updating service and profile settings uh updating images that are used using custom images how you can scale it up custom size your images so basically and high value basically all the hal commands that you typically use to manage the component sizing and high availability of the components uh you can manage them through a spinnaker or through the health chart so i will show how you can do it and then you know any additional health configurations that you want to do that you can do from here so i'll show you some of those things and again a walk through all the a few other configuration items that are there like storage and all that uh working along um you know if you want to work along with me um if you have a kubernetes cluster readily available i think you can try that as i uh walk through it uh a requirement that is there is that you know it's good to have a persistent volume available because um you know as you prob i'm assuming all of you guys are familiar with the spinnaker architecture i'm not covering that um at all here um you know it uses some storage right to store the pipelines runtime data and all that so it uses a redis and a menu and both out of the box and and those require storage and it the sunshine by default uses persistent volume to store this to access it it's nice to also have a load balancer available if not like i'll show it to you also how you can just access it locally but you know if you want to make it available for others to see it's it's good that you have a load balancer available in your kubernetes cluster now if you use any of the publicly available kubernetes clusters like you know eks gke or you know azure aks they do provide support for these out of the box and today i'm going to use actually a kubernetes cluster in google cloud gke to to show and tell now um whatever code i'm going to use or samples of it are available in this git ripper and a little while ago i posted that link also um in the also in the uh chat so you know if you want to follow along you can just click on that link and it will take you there with that let's go to this um github link so you know if you go to spinnaker summit 2020 there is the spinnaker um there are other stuff in there you can look at also in there which i use for some of my other sessions but you know today i'm going to you concentrate on the section where the content is in the spinnaker directory before going in there um there are a couple of things happening with help especially um repose publicly available ripples there was the stable ripper if you guys have been following or know about it that repo they're sort of not supporting and they're basically making it a distributed owner owned uh public reports right so um i checked yesterday this is still available where you can use the stable repo that's hosted in google cloud and so you can use that i believe november 13 they have been saying this will be you know discontinued and not available anymore and in fact what we did is we took it uh what were whatever was available in uh the public ripple and have replicated it in the official way to manage these repos going forward and we plan to work with the community to get uh this within the community role this really should be owned and hosted by the spinnaker community but till that happens and you want to play around you can use this together and i'll be using this one today to to run through it again this is exactly same replica of what's available as part of the stable spinnaker so let's go ahead and kick this off like i'm showing it to you here and what we'll do is that if you don't have um persistent volume and or our load balancer mainly the persistent volume i'll show you how you can turn it off and work without a persistent volume but let's first go ahead and run through this okay so if i go ahead here and look at how repo list you can see that i have added this repo already here uh so if you do this run this command help repo and give it a name you know i gave the name spinnaker and this url it's basically added as a repo available uh to help right over here right to install it with all the defaults nothing else just take the defaults and install it for me i'm going to basically run this command help install give it a name space i want to install it to give it a name and say use install the charge spinnaker in the repost spinnaker a lot of words spinnaker word in here right but just wanted to point out maybe i'll go and update this so that it's clear what each parameter are so this is the name space in which to deploy this is the name of your helm install uh this is the repo that you added and this is the name of the chart within variable right that's what we're going to install let's copy this in here oops okay there you go and let's look at our name space cube ctl get bars minus minus and spin occur and you know you don't see any spinnaker i mean uh running in there there's some remnants from my past installs um and you know trying things out before this session to make sure everything is working and all that so what i just did i just kicked it off right it's helm installed in the namespace spinnaker spinnaker right like i said it should take less than 10 minutes and we should have a running and a working spinnaker available with that uh let's go look at a few things while this is running like i said if you don't have um storage or if you don't have a load balancer mainly the storage uh what can you do i have lost my cursor and here i got it back so this is a local copy of that gate ripper that you just saw and this is that spinnaker directory and this is the values.yammer uh if you look at it um it is going to install a spinnaker version 122. doctor okay and it's going to pull in the images hallyard image from the public open source location where the hillyard image is available and it's going to use the version 132. i'm sure all of you guys already know if i go to spinnaker xpi versions it will tell you you know what are the versions that's available that you can use okay these are the current stable versions and also if i say hall yard versions i can go ahead here and i can look at the versions available so you know 13 days ago they release 140 but i'm using 138.00 now if you don't have persistent volume right and you want to try it out like you're using say local menu on your local laptop and you want to try this one out what you can do is ahead of time create some persistent volumes giving them around 10 gb each or you know you can turn persistence off uh set it to false then it will not try to create any persistent uh volumes for hallyard here uh you know i don't probably need to say uh that if you uh do not set the persistent to true you basically if the halyard part goes away all the config changes that you made would be lost right now the good thing about using helm and all that is that if you run it again you will get it back but you know if the helm part goes away and comes back up any of the changes made those are lost though so so that's the risk so you run into but if you want to just play around with that and see how it works and you don't have a persistent volume and you don't want to go ahead and create a persistent volume uh outside and make it available you can just turn it off the two other places where you need to turn persistence off in the default setup is if i go all the way down and let me search for mania uh it menu is where it's where spinnaker uses to store all the pipeline information right and there is this uh settings over here persistent enabled false so you can turn it up by default it is enabled true so you don't need to do it but if you don't want to use persistence with menial you can basically uncomment these lines basically these two lines if you're uncommon and run then menio will not use the persistence in there and the other place is redis which is right on top of this i believe it is right here uh redis again the same thing if you don't want persistence or don't have it you can just go ahead and turn it off right over here now while i was talking about this i noticed the hand chart came back now let's look at the pods that got created uh cube ctl get parts minus and spinnaker uh you can see the pods are coming up now all of them are not ready yet a few of them are still getting uh ready so uh once they are ready i will uh do it okay now once it finished you will see it gives you some information is to like you know how you can access uh the spinnaker uh locally basically and that's how we'll do it basically you expose the deck part and cue forward port forward to it and then you port forward uh to the gate part deck and gate and that's all you need uh to be able to access and like they say that you know if you want to customize it nothing stopping you from connecting to spinnaker uh hallyard pod and then run the hal commands if you have been used to running the hull commands but uh you know my recommendation always is that if you go run hal commands in there you lose them really uh if it goes away or you lose your persistent volume or anything like that best is to always come and edit this values.yaml and add them in this file right so they're persistent and all the commands that you can run through hal are i don't put up basically run them multiple times it doesn't have any uh side effects basically right so you can manage those values.yaml in your get source control it approval so you exactly know what's getting deployed and manage that deployment so my suggestion of a good practice is that don't go to hall yard and run the commands but update the values.yaml to make the changes right let's look at it if all the pods are ready or not they are ready and actually uh we did pretty good um less than five minutes or let me take the initial one when the hell yard came up um in about six and a half minutes we have our spinner cut up and running right uh it probably took a little less because the images were downloaded but you know you can probably give a minute maybe maximum to this for it to download the images otherwise you know seven and a half to eight minutes in a brand new cluster you can have spinnaker up and run let's access this like they say uh i'm going to just grab this thing and after i've done done this i will basically you know show you how you can expose this as a service uh so let's get the deckpod uh url and let's um put forward this thing let me run this in the background so that i can run multiple uh of these things over here so i have a deck being port forwarded locally and let me then do this with the gate pod so now i know what my gate board is and let's put forward the gate part to the localhost so i have both of the these being port forwarded right now let me go to the browser window and connect to localhost colon 9000 there you go i have my spinnaker up and running uh in less than 10 minutes that i have a spinnaker up and running that i can start playing around with it has installed the version 122.2 122.2 and i can start go ahead and create an application um you know i'll not do anything else just let it create um you know this is not about spinnaker i'm assuming you guys know spinnaker already or if not you know there are a lot of other sessions going around uh where you could actually go ahead and learn other features of spinnaker uh and um there are a couple of you know blogs i've done earlier on these things so you know you guys can find it but uh just to let you know that hey i have a spinnaker up and running that i can actually um uh use and run so let me create a uh pipeline here first pipeline you know i will just make it a simple uh pipeline where i have a wait stage uh let's wait it for five seconds you know traditional thing let's add one more stage which is a manual judgment easy it doesn't require much things in there which is triggered after the wait stage will ask for an input proceed whatever just some input that i can give right which i can use for downstream processing and let's save the changes so you have a pipeline ready over here uh let's try to run this manually right it triggered off it will wait for five seconds over here and then it's waiting for the manual judgment let's say proceed and i say continue and voila the pipeline is done so you know all within 10 minutes i have a spinnaker up and running uh created my first hello world type pipeline and you know it's ready okay let's uh get out of these things i had the other one running let me kill that thing so what you saw is how easily i was able to bring up a spinnaker instance using help all i needed was an available kubernetes instance let's uh talk about the next thing which a lot of people ask about and i'm going to go here is how do i make it available i i access accessed it from my local laptop right now i want to make it available to tom everyone's virtual these days right who's working from home and i want to make this available to tom how do i do that so what you can do is basically create services so if you go to that same github repo there's a file called spinservices.yaml right what it does is that we basically create two node ports uh for the spin gate service and then what we do is we create an ingress okay now oh if you are not looking for authentication you could always uh just create a load balancer service uh for spindex and it would work right like i'm here if i actually if you guys know this if i say gate slash applications it will return me my application i did something wrong here oh i did shut these both of these down so let's do the port forward to 9000 all over again notice i just pulled forward it to 9000 right i didn't do too bad um so you can see that it just found and returned the applications that are available i did not expose gate here so what it means is that the default uh spinnaker installation through helm charts what it does it it routes gate through this path gate and in halyard it's configured if you know the overrides it is configured to route a gate request through slash gate so it works so what it means is that if you don't care about authentication you could just create a load balancer service to deck and you are done but if you want to do authentication specifically if you want to use ldap what you should do is you need to create an ingress right so you create an ingress where everything is routed to deck except these two extra ones slash login slash auth which are routed to spin uh gate services which i have created as note port services because these are not exposed externally externally the access is through ingress so let's uh run this file and again if you have any questions what i'm talking about and you know any other comments uh feel free to um you know uh ask in the chat and i would be happy to answer them so cube ctl apply minus f uh is the spin services.tml and mine is in the namespace vinegar so it's in spanner [Music] looks like i had an ingress from an earlier time already in there so it's going to reuse that ingress so let's look at cube ctl get ingress minus n finicker there is an ip address and i think everything should be ready to go let's see if we can hit this or we need more time so that's the ip address and i go that ip address enter uh so the server is not yet ready yet uh it's probably because let's give it a few a few more minutes it's probably not yet ready so we'll come back and look at this let's talk about a few other things and then we'll see if the ingress stabilized over here so we talked about what all enhancements or what we can do so if you go look at that same uh gate triple i have this thing called values advanced dot yama that shows all the other enhancements that you can do right uh first thing which i'm going to talk about uh is not this section but other sections uh here uh if you have been making changes to your configurations to your profile configs like updating gate local or cloud driver local you can basically make them available right here right additional profile config maps and you just put them in here now you don't need to do this actually right in this file if you know what it is you can go ahead and create config maps and uh just see what that config map is right you could do that but otherwise you can just define them right over here so this is where you define the profiles and this is where you define the service settings now this thing is quite um confusing sometimes um as to you know what stuff i put in profile configs and what i put it in service settings the way i basically try to look at this is things that are needed by halyard to bring up the service goes into additional service settings things that are used by the service at runtime goes into profile config maps right so gate local if you want to make changes to gate local uh to uh say enable this um you know uh crop to enable basically this remote protocol header http server port mainly needed for course uh you just basically put this in or you want to enable logging at the cloud driver local uh so you can basically set the logging level so anything that you would set in a service you can just set it over here and then do and how uh deploy that will take care of things that you do at a at a service level which halyard uses like providing the image to be used right you can provide that here again these are samples that you can use to and edit it one thing which i've seen i need quite often is this environment variable to define the menial access keys in some environments um these are not set by halyard correctly and so you have to inject those uh in into the front 50 pod so you could do that so that's where you control the service settings and the profile settings if you want to use custom bom how do you do that you basically define it over here otherwise you'd be going and creating these files in the file system in hallyard and halyard would pick it up right but you don't want to do that in this case you want to maintain all of these in sort of source control and manager this is where you define it in in the values.yaml in the bom sections or if you have specific service related configurations you add them in the service configs here okay um you can control the hallyard pod how much memory cpu it needs by default you know it's uh not limited but if you want to control the memory uh and cpu you specify the address limits over here if you want to use the custom search for the halyard part that's you that's what you defined here now if you want to use this for other parts you basically run hell commands to do it and that you would introduce in the um up there and i'll talk about it here out of the box you can also provide docker registry that your account should be using like if you have a custom or private docker registries you can basically define them right over here like say if you have a private gcr repo you here the username which should be json key and your service account like the json file that you download when you authenticate against a gcr repo you add that over here so you can actually configure docker registries right over here um kubernetes uh two things here you know when i installed it i installed spinnaker in the same uh kubernetes from which i'm basically running it but nothing stopping you from uh installing um spinnaker in another kubernetes right or let me put it differently uh have your uh spinnaker deployed to a different kubernetes so you can define that right over here right uh what you provide is the cube config file uh it's either in you basically load the cube config file in a secret and uh what secret it is and in the cube config file you say what context to use and uh what should be the deployment context in which spinnaker would be deployed and what's the context used by uh for deployment by spinach and you can have all these other settings that you typically do when you add a kubernetes account in cloud driver you can define the name spaces or the omit name spaces and even like you know does it should it show only the kubernetes managed systems or everything else again like you know that either name spaces or amit name spaces can be provided you cannot provide both so if you want namespaces you should uh enable uh uncomment these and comment this out or if you want to amid name spaces just list out the ones that you want you can also define ingress um i i used a file outside of it to do it but if you have a host name and all that available already what host name to use you can use ingress right over here um basically your ingress should be uh or your host name that you want to use for the ingress is set up then you can use this to set it up when you're deploying if you want to deploy like especially in kubernetes right you want to have node selectors you want to specify where the spinnaker is deployed or some tolerance you want to define you can basically define them here right again these are kubernetes definitions that you include over here so so like you can and then we walk through the redis uh you can also use an external redis we did not give an external redis so it just installed a redis by itself same thing with menu uh we asked it to use menu over here and you know you just don't need to use menu if you want to use s3 google cloud storage uh azure storage all that you can define right over here to be used right so you can see this helm chart provides you a lot of flexibility in setting up your uh production uh spent okay let's see we i want to run through this and see if this one actually came back there you go right so it's stabilized it came back and here we have our applications this is my spinnaker summit 2020 application uh with my pipeline that i created right so this is available as uh i wanted wanted to go to the pipelines yeah this is the spinnaker summit applications did i lose my pipeline maybe i did again i'm not going to debug this here just trying to show you how it works and all that um but this is how you basically bring up a spinnaker now let's uh show something that we can see quickly is uh talk about a few other things here uh if you are in a restricted environment where you need to use service accounts to run spinnaker and all that you can define them right over here in the in the values.yaml file if you have any questions on on the values.yaml what i can do or hey can i do this or can i do that feel free to ask me what i did is really walk through the values.tml um you know so yes you could have always gone through the values.yaml but you know sometimes you know trying to relate it uh to spinnaker is what i think is missing in values.tml that's what i try to do provide that input to you uh now now let's come to the final thing what if i want to run some commands that is not available in values.yaml right i want to run a bunch of highly out commands you can do that actually so it will actually run these halyard commands before it will do a hal deploy apply and all the helium commands that you want to run you can just type it in here as part of additional scripts again the syntax i'll not go through exactly there's some new answers in how you set this up you can always go look at values advanced and look at it but main thing is that you can give that this a command in this way help config deploy component sizing deck edit so basically i'm saying that run two instances of deck run two instances of front fifty that's what i'm saying here right and i can add whatever any any how config commands right over here and they would be applied right let's run deployment with this one and you know see if we have multiple deck modes running or not so if i go look at it cube ctl get parts minus n spin occur so we have actually it probably ran that file i have two decks already so i probably ran that already so let's do something else here i'm going to remove this one out now and reduce this to let's say three instances from that and i do what do i do is held list minus and spinnaker and this is my spinnaker running so upgrade uh in the namespace spinnaker uh the release pinnacle sbi spinnaker uh from the repo spinnaker uh the charts can occur lots of spinnaker here um and i want to use that uh advanced right values advanced.yaml and let me trigger this off what i'm expecting to see is that once it finishes and comes back i expect to see there are three uh deck parts running over here right that's the change that i made over here let's wait for that while it is doing uh let me talk a little bit about another thing that i hear people talk about quite a lot uh is this uh um how do i uh any configure ldap right i know where my ldap is again like i talked about you can always include this in this section right here over here so to enable ldap what you do is you configure security and you basically enable ldap and then configure ldap right so you need to give the user the end pattern you need to know and the url to your ldap right and there are a lot of other parameters specifically like if your ldap needs authentication to connect to it like you can give the manager dn and all that also as parameters in there you can look at the hal config security authen um authena command to see how that works and then you have to override it the base url for for your ui and what would be your deck host name here it would be the same ingress that we provided right it will be the same ingress ip address or if you have a host name for there you can use that and once you do that you do a hal deploy apply now if you're doing it through this file here you don't need to do the hal diplo apply right you just do these three commands uh right over here and it's it would do it actually enable authentication with your halyard uh it's running over there so let me just add this over here and you know if you have a little bit of a time uh we will try to uh see this uh over here this should be see and and like i said that you are a little bit this guy here i didn't do not need an extra level of http uh but that's about it so let me save this thing here and see if we can enable um authentication to my uh environment right from here now i'm using an open ldap that's running in the same name space it's not in default it's in main space finicker that's the dn pattern uh basically a vanilla install of an open ldap um that's what i don't give it to you so let me save that and i notice that this is finished over here right so let's look at the number of pods running over here uh there you go um you can see there are five deck running but you would notice that two of them are terminating right so i increase the number of deck parts running to three so let me just trigger off another one over here and um you know and see if it gets enabled with um open ldap so i've again uh saved that file run ran it uh i don't have anything else up to show uh if there any questions i would be happy to take it and this takes about um how much time uh a hundred or or maybe seconds so which is about like two minutes two and a half minutes i'll hang around for another two two and a half minutes and see we were able to enable authentication with our uh spinnaker and um i'd be happy to answer any questions if you have any while doing that let me just summarize it again if you have any questions i'm available and reachable at uh on my email nirmalia opsamex dot io i'm also in the spinnaker slack channel at nermaliasen and if you have any questions feel free to reach out if you have any questions now i'd be happy to answer it and just to recap we saw how you can using home within 10 minutes bring up a spinnaker instance and start playing around with it uh in in in google in a kubernetes environment and then all the additional uh configurations that you can do through through helm that helm exposes and anything that help doesn't expose you can actually just run those held commands and helm will actually run those halyard commands and help won't actually run those commands through so um let's see over here this is still running sometimes while it's still running this gets done let me try to reload and see if authentication is enabled or not yet not yet we'll give it a minute or so here i'm impatient here so let me just um command t uh let's do smaller window command actually i think i lost my other window i don't want to do that let me go back here let's let it finish again like i said if you have any questions i'd be happy to answer them now one thing that can happen if i did not given the values correctly it will wait for it and then let me time out but um i'm thinking it should be good authentication not enabled it's standard hopefully it's not your demo blues let's see while this is finishing if there's anything else i'd like to point out here uh that would be useful uh yeah the other thing is that you know you don't need to type it in here you can create a config this what it would do is it would go ahead and create a config map uh in in the kubernetes and pass it on to halyard and halyard would use that so you know you could actually just create that config map outside of this and give the config map name and and use that right you could do that also uh and and have it use that config map uh you don't need to actually type it in here and then that's a pattern used at a lot of places also where you know you can actually just give it a config map where you have actually created the values ahead of time okay um let's see it hasn't finished it yet and i think it is more than the two minutes it took in the earlier case okay definitely it timed out it took more than it should have taken and something did not work and gate did not restart and we'll just look at uh want quickly try to look at look at what went wrong this is the task that runs it that is still running cube ctrl let's do the minus p minus um some some parameter i provided probably was incorrect and that's basically failing on that one uh let me not do an n o minus b uh so there was an entry uh oh i probably made a mistake i just said how i did not give a command something went wrong over there it tells me over there i missed out the hal config that's why it failed in there but you know again that probably brings up a very good point uh what happens if it doesn't go wrong so that's actually sometimes it's a good thing and that's where i'm going to end today is if you look at the parts there would be a part that's called install using have what it does is does all the prep work on your behalf and so if something doesn't work correctly like you know this is not completed like you saw see over here it's still running that means something went wrong in the configuration and if you can if you look at the logs of this it will tell you what went wrong so that's what we did we looked at the logs of this one and it clearly tells me that you know it expected a command here uh and it got security um so i did not tell it that how conflict security and so that's why this failed right so you know although i was not planning uh to talk about how to debug but because of the mistake i made over here not giving the config it actually gave me an error as to what went wrong with that i see no questions uh i'd like to thank you uh for your time uh thank you for listening to me if you have any suggestions how i can improve this or if you have any questions um feel free to reach out to me at nermalia.opsamix.io or i'm on the spinach slack channel at miramalia