How to Discover and Expose ECS Services Automatically in a Live ECS Cluster Using Traefik

good morning everybody it is 9:00 a.m. and I am in San Francisco and I hope that you are doing very well today I'm sitting here with my coffee and I'm getting excited to hear our presenter today talk to you about some special stuffs so first let me tell you a little bit about Who I am contain us and we'll also talk about what the online meetups are so my name is Patricia Dugan I am the head of community and I represent our beautiful traffic community and I work for containers and we make a traffic mesh and great open source products to make infrastructure boring meaning that we want your life to be super super easy and fun so what we have today as a traffic online Meetup and traffic online meetups are expose zuv ways companies and developers are using traffic to solve interesting technical challenges and these are demo rich technical really yummy for people and they live on the YouTube channel so you can watch them over and over or watch them at any time let's see a little bit of housekeeping as I say we are hiring at contain us we like developers we're looking for technical writers developer advocates and all sorts of roles so go to our website and go to backslash careers and check it out and if you want to see what other virtual events we have coming up go to back slash events and see where you can find me some of our team and other guest speakers talking about how they're using traffic okay what do we have today well this one's going to be interesting today we have my friend Vikas Kumar dialing in from Australia he is a senior cloud consultant at Rackspace and he is going to present to us how to discover and expose ECS services automatically in a live ECS cluster using traffic this is the first of its type for our online meetups so we're going to learn something new today and with that I would ask that you ask all of your questions in the chat box and we'll answer them at the end and so now I'm going to go mute hand the mic over to our friend because and wish you a amazing show and hope to see you in the future at other ones thank you for coming today have a good have a good time with us bye bye hey guys thanks for joining good morning afternoon or evening depending on your time zone so I hope you guys are doing well and staying home and helping common to avoid spreading the virus so as participation my name is Lucas I am from India now living in Australia I like space as a scenic route consultant and help customers to architect and deploy the cloud infrastructure I come from a solid into a background started my career as a UNIX is tremendous waiter and dive into cloud environments in 2013 and have been doing them up see stuff even before they was become a buzzword mmm I came to know about traffic about two and a half years ago and have been using both traffic wasn't and wasn't you I mean using traffic wasn't one and two you know in quite a lot of my personal official and freelancing products in a variety of use cases cumin it is a WCS you know file back in stock just a few providers to name it is my go to reverse proxy for a you know or having this controller in Cuba it is unless I have a very special need to do something something really nice so again today I would be talking about traffic wasn't one in Amazon Elastic container service or simply put as easy case environment now so so before I move further as you might have noticed I have a heavy accent so just in case you do not catch what I have said feel free to raise a hand in a chat window or just write it up comment on the chat window I would be monitoring that and I would be happy to you know repeat or rephrase my sentence so in this demo what you will learn so you will learn about traffic capabilities to route traffic based on HTTP headers and darker labels and as a bonus you might want to be you know interested in a few things about Cara for Quran and makefile so this demo is obviously about traffic and you know but if you have any questions about any other tool the technologies used in this demo feel free to hit me on LinkedIn Twitter github wherever you like and I would and yeah that should be all so I would be following a github repository in this demo demonstration so I would miss this Paula tree link in the chat window here you go now now I have written a small we'd be page you know which I want to follow in this demo and if in case you are trying to follow the same just make sure you use an Sydney region because that's what where I am and that's why we have coded so I'm not no I'm not putting a lot of efforts in terms of enabling it in terms of region this at this stage and if in case you have you see an area for improvement please feel free to raise PRS I'm happy to welcome all the other welcome I've tried to keep things simple in terms of you know spinning up the environments and even ripping it off just because you know I don't want you to our third point any leftovers you know it lying around and you and will you so let's get started okay so once first of all you know why I have put this in the first place what problem I'm trying to solve it perfect and you know why why we hit why we at the first place so you know one of my customers wanted to use ECS and obviously you know I my end goal was to make it make the use the end users experience as good as possible or as easy as possible [Music] yep so ya know don't worry you don't have to see my pretty face for a long time I would be sharing my screen don't worry about that that's for s go buddy it's okay now okay so I so what I was talking about is why I have put this together in the first place so whenever my customer came to me and you know he wanted to use EEZs and you know there was proxies and things like that now the scenario I had was he had a cute few question applications running inside docker containers in the shape of easiest services and tax temptations now the next logical step is to expose those services you know so you have a HTTP or an HTTP endpoint of it which further hits turns to the containers now what's the default option you have the default option what we have is an ALP an application load balancer what you know what what aw sacraments now in terms of alb there were a few downsides and I wanted to you know as I said I wanted to make it as easy as possible for the end users or developers so maybe I should start sharing my screen now give me a sec I'll start in my screen yep so just let me know if in case my spleen is visible to you guys okay I hope that screen sharing is working fine okay so yep so why not why not just publish a load balancer just in alb in case you google around this this is I think the probably the first or second search the second link you will find an application load balancer now why not an efficient lost but load balancer the capital balancer is good it does the job but again I wanted for the end users to have to not you know mess up with my television which I have as an AC segment and I wanted to make their job as easy as possible when I say easy they should they should not be you know messing about application load balancer rules or you know target groups or secretary groups in alws all they should care about is really their application or the containers so I you know I googled around and probably found a few things but they were not nothing which I could readily use so that's why you know I came up with this solution and no I try to use traffic in this case so so how does traffic helps us so if in case you know I just want to point it over here so if in case you are just using an alb how would you normally do it you would have an end user then you would have some application load balancer and some certain use over here in this case for example if in case you have n services you have ten rules over here and you would have ten target groups ten security groups and you know number of number of basically for every service you would have load balancer rule and a target group now now how does traffic helps me here over here so traffic basically you know there's a lot of things for me it just removes the necessity to have multiple application load balancer rules and multiple targets just keep the elements simple a lot of this architecture would look at simulator if in case you're familiar with cumulative it would be it would like an in like an ingress controller so traffic over here over here you know routes as per the host headers now now this is this is what traffic Helsby now going down the slide there are few other things which would traffic natively provides you as well now if in case you are familiar with traffic you might see that you know wasn't to is latest one and reading wasn't one now why is that I'm using wasn't burnt one because you know I wasn't - still doesn't have and easiest provider and you know and wasn't one it's still being developed so it's okay to use it but for other all the purpose I would strongly recommend to use was into but for easiest rather than one is one is good enough okay so if in case you want to follow on you know I would be giving a live demo if in case you are following along with me just make sure you have these tools in salt this is what my repository uses and but you know my again I have framed to the poetry in such a way so that you can even if you know after this demo you can easily smell up infra and until it don't run in your text except the exact so if in case you need if in case you guys need need time to download these binaries or these these software's I'm happy to wait for a couple of minutes just let me know in the chat window that would be good all in the QA session whatever to whatever you feel like can't see anything okay I'll probably move on good so what what we have so basically what I will do is I will spin up any CCS I will slip an ECA sintra with some sample services and sample task I will be spinning spinning up a new brand new V PC in order to do this task if in case you haven't it doesn't really matter because this V PC will be will be a new end and you can still use in an existing database environments okay let's get started I will I'll put this put this window in the upper part okay and then I will start with the so again if in case you want to start with you can still clone the repository the repository is in the chat window this is the repository and we can get trolling are you guys able to see my command line I just let me know if in case you want me to make the font bigger I hope it's okay cool okay so before I start I would like to give a very very brief into infrared action about you know okay I have something a chat Hey okay good thanks bye teacher good so I would be just giving a very very brief introduction about my you know how I pictured the cold so this is obviously I'm using telephone to spin up resources and unlock a file with traffic here it is I'm just you know I've given a custom entry point to docker so no if in case you're familiar with docker and Depot it is something which gets executed once the container starts running so in this custom endpoint I am passing the traffic configuration which looks like this so traffic in this case you know it's pretty easy so if in case you have worked with the reverse proxies in term like engine X OHA foxy one thing you notice about traffic contraction is it's very very simple and they know it's very minimalistic so you don't really have to manage a lot of virtual hosts all virtual hosts are managed to drop your labels I'll show you 101 see progress to the in the demo so this is basically these are the cluster host and a domain and my my region obviously the thing is this is all you need and I'm using ACS back this is good what perfect all the DC is provided so this is the easiest provider which I am using and then I have a file folder I'll show you how it looks like in a moment and in the entry point I am just replacing some values because I want I didn't want it to hard code these values because you know going forward you might want to use the same image somewhere else in your environment so it would be you know easy to just reuse my or you use what I have done so far you don't have to know the fee I think the whole thing again ok so this is what it will use for the docker images and this will get pushed to your own ECR and then here is where armitage definitions for my for easiest us and then easiest terraformer you I'm not diving into it now but again this is tara forms and all docker if in case you have any questions feel free to hit back on twitter linkedin wherever you want okay now all the commands all the telephone commands and all the commands in you know make file just to you know make things simple for for you in order to if in case you play around with this in your own environment if in case you're following around with me in this demo just feel free to do use use the same make files I have just wrapped it around so that it's easy for you guys and over here this is the same readme I was just opening it so that I can see the commands what I have so I am exporting my AWS profile which is it has full permissions to to manage the easiest services and easier just in case you if in case you're not sure what to use I would and I would use I would come into use you know a demo demo it abuse account or non-potable second with a new set of abilities if you're not sure what what we'll use to use and this is my obviously my my Sydney region okay so first of all I am just building a docker image and pushing it to ECR so that you know that in each can be used in my easiest definitions so this would create a nice this would tag an image push it to an ECR and wraps it and it's easy to keep here for me just in case I need to login I don't have to log in but just in case I need to log in then then I have a make just initializing terraform over here I think you're familiar with terraform you might know that you might need to initialize it first in order to download all the telephone providers and all the things so you can do that and that's that then I am applying so the apply so what is what this is going to do is basically build up in easiest cluster for me and with some definitions with just two sample services and obviously it's traffic total in total there are three services one for traffic and two for to sample services now this will take I think about three to four minutes I think is I'm not wrong yeah you now I have so again these are using the start options which I have over here these are not some normal JSON files very simple to use very easy I'm reading traffic as has been a diamond mode what did what does that mean the earth mean is that you know on each instance of your on each instance of your easiest cluster traffic would would run on a host put on put a tea in this table obviously that is not that is not production ready this this whole demo was not not not meant to be production ready but this will give you a good overview of how you can use traffic so in real life scenario you won't be using port 80 or you would be using port 80 and handling SSL offloading load balancer it all depends upon upon your environment and all the services are using the host port 0 this is pretty important because we don't want to eat the top when you use a host port 0 in easiest efficient tasks what that means is that basically you are running your tasks on a random boot so what does that allow that allows you to run multiple containers for example you have an application container which you want to have five instances of your application so you if would be running on five containers so now now you know you can't use the same host port for more than one application so this will just assign a random Hospital and same for this this example I'm not using a random was put over here the only difference between the these two applications are one uses an environment variable show IP faults and other uses that doesn't have any environment variable so these are just format from sample images if in case you want to look at the sample images I would show you what here it looks like this this is a sample image I'm using site counter image it basically gives you number of hits how many times you have hit an application in this case and it also shows you initially sure use the IP address of the container as well this just I use it for from a local testing a lot you know when I have to do something nothing very fancy about here it just uses it's written in goal on and they just use this talk basic stuff okay let me see whether my my container is ready okay so this is still being created okay in the meantime I want to show you how the how the flow looks like once we this okay so how would you normally you know if this is a double-double pattern so what do you normally do if in case you want to have a common load balancer for multiple applications what you end up doing so normally what you end up doing is creating C names or our output III alias for example if in case I want to host multiplication on this a pillow to answer for example abc.com and xyz.com now I would just create a cname in whatever your DNS provider you are using public or private I would just create a scene him to the DNS name of this efficient road balancer and then the traffic will you know I mean not a traffic in traffic a normal traffic HTTP packets will flow from a load balancer to a target group and then point it over you know in this in this case traffic docker container the backend services now this this service name over here is an easier service and so that is the normal CR you how you do English controllers or all a reverse proxy setup so this is the same same stuff I was followed okay that means double check okay so this is added so once you apply this it will give you an output dot txt make it bigger okay so so this will this will give you an output txt you know and it just gives you some sample commands to run again I have just made it simple for you guys so you can just copy by system on and it will work for you you can use curl or I uses something else over you tiny routine equal HTTP by just if it is just a wrapper a bit nicer I would say okay I think I would have to just update my traffic DNS over here and all I'm checking now is whether my traffic is healthy or not or its way to serve traffic in other words okay so if in case you're following along and you do a Perl or you know actually P whatever you want so if in case you see AC direction that means your traffic's healthy now I just want to check whether my services are healthy or not now all I have to do is use a different host shader which is in this case a site counter one yep site counter - okay okay so okay okay so this is this just reflects that my setup is good it's all working fine and now I'll be back a bit into it now what is what basically this demand is doing so I so if in case you any of you is not familiar with what this is three all I am going to all I'm hitting this this DNS endpoint which is of my alb and this is the host header I am passing this is the host HTTP header which is a common way to fake your you know fake your request and passing a horse trailer so that so that Venus so that they'll be thinks that that someone is requesting on for this particular website or a service how you want to point it out okay so coming back to the console and so this is my console I have a this cluster running and this again I said there are three services one for traffic one for and two other two for my applications now and then I have few tasks running again they were two containers in each task or in each service or two tasks any services so that's why you know you have five pipe tasks to each for application and one for traffic why one for traffic again because traffic is running in a diamond mode now these are just for instance running over here with with five tasks obviously you could you would have more but in this demo I'm just using one and that's pretty much it in terms of you know sorry services and tasks and then in the tough decisions you would see perfect-perfect parks technician and this is a JSON and these are the environment variable aren't passing I think ace you remember in the docker entry point I have been just replacing these values just to make it more customizable so that you know if you can even else once you use it all he has to you know parts of different environment variable in this this big JSON this is the traffic image I have been pushed in the first step to do my ECR that's that's it okay now I want to show you how traffic basically is routes the connections in other words how is traffic rocketing the creation based on horse-trader so if in case you you have worked with different reverse proxy like engine X or H a proxy you might have seen you know we killed a lot of virtual hosts over there and each time when you create a virtual host or when you modify what your host when you update a virtual host when you create a new one you have to reload your service so this was actually the one thing you know why I started using perfect because nothing actually took the piece overhead away from me I didn't have to reload my service I didn't have to you know configure my service it was just it traffic just sits there and you know watch whatever provider you have in this case ECS and just reload itself or just configures itself automatically you don't have to do do that so this was a huge plus for me and this was the main reason you know I started using traffic apart from that the they are quite a few other reasons as well I will just point one or two more one of them is a suffix than a very good API so you can if in case you know you you are you use a lot of APs it's very very good so in this case if in case I you know go to my it should be and endpoint and type API it will give you all the API endpoints I know whatever I have configured so far you can also make it should actually be called and configure those things so so these these things and having API I mean in my opinion opens a whole whole door you know for automation you can do a lot of automation between you when it something is API compatible and the other thing good thing about traffic is that it gives you a very neat UI it gives you a very good options just based on labels you can you know for example you want authentic Asian you want to know protect your site or your service with an HTTP password so it's very easy to do that just based on labels I would highly recommend to you know go to this particular URL and just look look at the different options you have I'll just face this you are in the chat window for everyone here you go so you know just see if you look or look at more options tons of options all there so you can have Kennedy deployments you have waited deployments basically everything almost everything about you what you want from from from a reverse proxy I would say and in this case also you know all I'm all I'm configuring is the column typing is my reverse proxy is true it's to dr. labels I'm not not hard coding anything anyway I'm just using dr. label stuff but that's actually all I'm doing and this is it says this is the whole magic I'll just make it by the phone bigger for you so this is this is all I'm doing traffic enable true what is my host side counter one might but what is the backend I'll show you what about back-end means in a moment what should i protocol HTTP so so basically that is all you know if in case I want to give this task to a to an end user or a developer this is all he has to come up with apart from his application this is all he has to come up with heed now it doesn't have to mess around with a lot of you know security groups or alb will be rules and all those things I will be free show you how a nail will do looks like I think if you have to do it manually or even in an automated fashion you have to manage the whole lifecycle of an of an application world balancer rules I'll just open this see this this is an example so in this alb I'm using a horse called one to the xyz.com XID abc.com now I don't have to do all these things in traffic just a one alb with a simple rule is enough I'll show you how it looks like easy to go to load balancers sorry you this is it this is just 1 lb our views are not created multiple bees because it's in some places you see the people create multiple lbs for every service or just one a lb per service but I don't which I don't think is good what I need to do so in this case we are just 1 lb with just one service with just one rule it's just set and forget you don't have to touch this again and again and I don't want you know end users or developers to touch it as an easy assignment or the cluster admin I don't want them to touch these rules I just want this to be set and forget thing that's why it just makes things easier so in this case I just have one rule which stays you know you receive traffic on put AP and forward the traffic to a target group target group is nothing but you know which basically contains something back and instances on certain puts in this case my back-end instance my target is is an instance which is running on port 80 and I have some health checks so in this case you know traffic actually listens on traffic just how true is it ok so traffic says whether I am able to serve traffic whether I'm healthy or not based on you know this is a ping pong check it just basically says traffic's healthy or not what is healthy or not that's a normal check normal eld your LD he'll check that's that's that that's pretty much I have to do I don't have to create a lot of security groups a lot of rules over here that is one thing I just as a certain forget again coming back to this particular scenario this is again I have to do only once and all the routing based on HTTP host headers are done over here again in the and in this case this is my service staff definition so traffic's just sees the so it starts definition and you know and passes the value now I'll show you how the how the UI looks like it's pretty pretty neat pretty neat should give me a schema sec I'll just copy my output dot txt my basically my DNS name which is this now what I'm doing is if now in I'm just faking my you know my browser to basically you know if this DNS it just just point to no traffic see later at this brightness taking this for sure in other words they're very simple words I'm just faking this fortunate now there are multiple ways to do it but I think this is the easiest way in order to show you guys so that is why I you know I just use it so this is a place the plug-in I think I don't want to waste time on Google so there's a plug-in called malt header try to use this plugin it's very simple to use and when you just use it you just have to you know put an entry over here to request for shader they were traffic and it would be good idea to lock your tab so that you know this host shadow is only used for what you have done so far okay so what we are seeing over here in this case we are seeing are two applications this is the front end obviously our back ends the front end maybe L should we get a big bigger okay it's a better now we have two front ends are one corresponding to each service again this is the Amazon services what you have traffic and service so so each service over here is the site one to one and side conduit to corresponds to one front end over here and then I am running two containers or two easiest tasks are the back end so this is the one-to-one mapping of a front end and a back end but this is how you know how traffic you UI looks and was on one and then I have a file backend which basically is traffic internal so whenever you know a traffic see la Piedad fun is basically points or suit this pack and put 8080 which is the container put running on each of the now you might see some some random posts over here so again this is an easiest thing so I am running the containers on port 0 which basically means a random port so whenever you know I deployed no containers I just use random ports over here now again so this is also you know why I wanted to use traffic rather than use an alb approach because now I don't need to worry about you know random ports because obviously you can't you know create all these ports manually don't have to you don't want to create this ports manually you and you don't have to even manage a traffic port in terms of alb you don't want to you don't need to do that so in this case I am just using this and in this case the weight you see is one wait weren't means you know you were just using perfectly this packet is just using a protocol a load balancing now I'll just give you quick run through this what what I mean by load balancing so so in this case I am pointing out to my side come to one service you see an IP of 10.3 over here that's a container IP now I point again I do see a full zero for IP if in case I'll do it again you'll see a 0 3 IP so basically it's like it's load balancing between 0 to 3 and 0 at 4 is the container is and why it's load balancing because my weight over here is 1 but it's by default and I have two containers now if in case I have more containers over here it will obviously you know load balancer between all this content automatically now now if in case you want to have greater load balancing or Kennedy deployments it's really easy to do with traffic all you have to just change your you know your token labels which again goes to your past editions goes to you over here these are your colobus you tweak these docker labels certain parameters and you know you will have weighted load balancing now doing this in traditional reverse proxies is a little bit of challenge if you have worked with in genetics a checkbox in all the things in terms of challenging I would say in terms of automatically pairing these stuffs you know because you don't reload insects all the time you don't want to restart Angeles all the time all the other things so traffic again traffic just sits over there watches these labels and applies it now and if in case you've worked with cube root of Cygnus controllers this this fundamental of working with labels and annotations would would sound familiar to you ok other thing what I wanted to show you was it has a health check so traffic by default why does this help dashboard you have so you have some photo force you have some two zero zero at 3:02 it is basically redirect and then you know some some bots are actually trying to hit it means so I'm betting a photo for so that's why because this don't worry about this for a force these are some bots on internet trying to it because I am using a public load balancer but open you know open security loops obviously that's not recommended in a production environment but this is I can do that ok so yep against you you have an API over here as well so whatever API I showed you what the CLI yellow over here you so I'm just using a JSON plug-in to to just you know have a well you need JSON so if in case you don't have this second you will look like this to you so this is a bit petty so again so whatever information I have on your dashboard it's on JSON so again this opens a whole door for two mission for you hold it into a photo mission for you that's it for the health checks so this is the your traffic wasn't I'm using the latest version till date 1.7.2 for that's pretty much so this is a pretty simple UI you know it just gives you a nice it over you of what's going wrong now if in case your your back-end is not healthy or something it will just show in grade or something and the good thing about uses in this UI that you don't have to reload it for example if in case now let's do it I think this I just delete one of the services this would automatically pop up on the UI and you would you would see that the services go on okay this thing I like about traffic is like it's dynamic I think if I am using a traditional reverse proxy I would have to do it I do have come up with some kind of automation to do this it's still doable I'm not saying it's not doable still doable the only thing is that you will have to come up with some automation but again you know if in case there's something already there why you have to do that okay so I am going to my clusters and tessellating a service so I'm deleting a side counter to service delete me now takes a while for a service to get sealed okay now this these services should should go away over here see but boom now you don't have a psycho to service now neither you have a back-end because that the task has gone and obviously it passed those things those labels are going that's why you don't have a service over here now if in case I go on my terminal and to make PF apply again what it will do it will it will provision that service again for me automatically that's that's what that's the telephone thing now once I have the service I I would automatically have you know my UI or traffic would start serving my let's wait for a few seconds for that task to pop up okay it's not there yet you okay so you have a cycle to service but there are no running tasks once you have tasks okay so now you have two running tasks and this should be available in my yep see hey so so this is this is this is a dynamic nature of traffic what you know what I really like and I think even if you this is not about creating a service even if you update a service with with more you know more tasks or more containers in other words I can still do that for example I would add a class over here I just change a number of tasks to two to three next step next step step a bit of service now I should have three tasks running instead of two really just waiting for the tasks you complete okay now I will any count of three now I go back see I have another task over here another task or another container and then this this container is also ready for load balancing and this is again this starts on a random port you don't have to worry about this easiest does it for you and you don't have to specify this port anywhere in a secretary power or something it just works just works um what I want you to show walls hmm I have already shown you the API the documentation just by the chance do you guys have any any questions so far just the spider question in the chat window or raise our hand all the QA below whatever you find easier absolutely be pretty sure particular solution for me okay hi to implement the health checks like a target okay sure so whenever in this case you know my health check in this case is I will go back to my ec2 tab and I go back to my load balancer so health checks is implemented on the target group in case of lbs so elby's listen on a particular port and just forget for forwarded traffic to up to a target group now they target group does the hell checks so in this case the health check is on the target room for two which is traffic now not traffic so this was the hell take of the traffic instance now in that how does Rafi does the backend health check it's really depends upon you know this is this is more of a talker he'll check that traffic he'll check question so in this case what what traffic C's is you know I can I am I able to you know hit this in this endpoint and get a response that is all Torchic is checking or or in the in other words is the container running LD now now if one of the doctor best practices is to define a health check for your container now when it really depends whether you use it or not in this case I not used it but but you know I really encourage you guys to use it dr. hill checks so whenever whenever a pentane is healthy you know traffic would would would serve traffic to that container otherwise it will just ignore it that that's basically how Bacchus basically got more of a doctor he'll check other of traffic he'll check so traffic in this case we'll only show containers which are running and which are LD now whether the application is running healthy or not that's merely depends upon how you design your you know your containers or how you expose metrics to your containers they help the hell check matrix does it answer your question cool thanks any other questions so far okay I'll take that I don't know mmm okay now I think that's that's pretty much I want you to cover so you know and I can't emphasize enough that if in case you don't you once I'm hope you're following my my instructions on your clusters if in case you are you're following it just make sure you you know you delete your question I have just been again I've made it simple for you guys just one this may come and make nuke and it will just remove all the resources what you have created so far yeah and you know finally I'm saying you know I have used traffic traffic a lot in my in my official and personal cipher tricks it's one of my go to reverse proxies on cumulative static files traffic we do if in case you come back of static we do that's a compute overall of traffic people and it's equally amazing I would definitely recommend you to use traffic veto if in case you're using various clusters it works just works beautifully maybe I'll give a bit of on some time on that as well that's pretty much I have to share if in case you have any questions of our even after this demo you are facing some issues or something just feel free to you know again people to fish me out to Twitter or github or big thing I would be happy to answer to answer those now I'm just relating my resources just make you so this will just remove whatever whatever I've created so far that's it I hope you guys enjoyed and your Twitter handle into the chat box I know not everyone's on Twitter but it's a great place to have real-time engagement with developers so it's veikkaus underscore zero to seven I think that was a wonderful session and you know it's very modern we don't have to fill up the hour it's okay very solid thank you to everyone for joining us and for your questions Benjamin really appreciate it and so what gonna happen now this this will go be turned into a YouTube video and I will email that to you and and then you can leave comments and questions for us if you have more later I also encourage you to please check out our events page we have several meetups coming up in the next month and two very interesting cross-section of people and companies large companies the next person is or the next one is actually three community members Turkey Switzerland and Poland all being representative of traffic ambassadors and we're going to talk about migrating to two point X so please join us for that and I think you on behalf of traffic community and contain us as a team for your time PICUs and really appreciated the session so thank you yeah okay cool thanks for your time thanks for your time everyone and now we do have a fist bump boom okay wrapping up see you later bye everyone thank you guys thanks

Channel: Traefik Labs

Views: 1,926

Keywords: docker, kubernetes, microservices, golang, traefik

Id: mYYwXHKjayo

Length: 44min 45sec (2685 seconds)

Published: Wed Apr 22 2020