Cloud Native Computing Foundation KubeCon + CloudNativeCon North America 2021, 30 Oct 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: A Deep Dive Into 5 years of Minikube - Medya Ghazizadeh, Google

Description

Don’t miss out! Join us at our next event: KubeCon + CloudNativeCon Europe 2022 in Valencia, Spain from May 17-20. Learn more at https://kubecon.io The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects.

A Deep Dive Into 5 years of Minikube - Medya Ghazizadeh, Google

A deep dive into minikube project on architecture, project maintenance and the future, 5 years of minikube, learn about the architecture, challenges and the future of the project. including topics on the future of the VM drivers vs Container drivers. future of different runtimes. and how minikube got improved the benchmarking metrics on minikube’s performance on starting a cluster, cpu usage and image build. also learn about the process of maintaining a project that has been around for more than 5 years. and then end share top tips and ticks on minkube that you might not know about.

A

All right uh good afternoon, everybody, my name, is media ghazi zadeh, I'm one of the maintainers for mini cube, I'm a technical lead manager at google, and I have the honor of being leading the mini cube team, the past few years past two and a half years, and then here in this talk, I want to share with you an overview of the maintaining mini cube in the past five years and also I'm going to share 20 pro tips of using mini cube and also some lessons. We have learned in maintaining an open source project.

A

I also have with me anders birkeland from yotoburg sweden he's online uh he'll be joining us in spirit and online lecture.

A

You can follow me on github or twitter. I don't have that many followers, but I will only speak about minicube and coding in this twitter media dev, uh I'm curious by the show of hands who has used minicube here.

A

Okay, that is almost everybody and another question who has known or used kubernetes more than five years.

A

Almost nobody, so I think minicube is older than all of us, including myself, in using kubernetes. I myself did not use mini cube. um I first used mini giveaway four years ago, so I think when mini cuba started, I did not even know about kubernetes at that time. The previous maintainers for mini cube uh dan lawrence started it and then thomas stromberg and matt earth, but the current maintainers up in the cube the people behind the mini cube.

A

Emojis are these amazing people under sherry, falgamal, uh stephen powell and frederik orokovic, and we are from diverse sort of countries and companies.

A

We have three googlers and uh two non-googlers uh in europe, asia, south america and north america and the middle east. So here these are the emojis behind the faces behind the names behind the emojis.

A

Here are some interesting stats about mini cube, mini cube has more than 685 contributors, which 16 of them have contributed more than 100 commits like a very active commit developers. You have about 3 million github downloads, which does not include our main method of downloading, which is gcs buckets, but still an amazing number of downloads.

A

You have a 90 csat score according to our surveys, and we have had 125 releases in the past five years. 125 times people released mini cube. That's really amazing. To me uh this is a graph of lines of code of minicum in the past five years, your 1.6 million lines of code for this project, uh which is amazing that includes the vendor.

A

If you do not count the vendors about 550, 8, 000 lines of code for minicube, uh by the way, this method that we chose to uh count the line of code, it does not include the comments and does not include the the docs. It only includes only the go code that we have written for minicube, uh an overview of mini cube, something which is unique about mini cube that no other local kubernetes tool could claim. That is a diverse set of uh technologies that you could use to start a local kubernetes on your laptop.

A

So whether you are a person who uses docker or if you are a person who uses kvm or if you are using vmware or parallels or even if you don't want to use any virtualization technology, you just want a bare metal non-driver.

A

We have all kind of solutions for you, so minicube is not bound to a specific driver.

A

It's it's all kind of virtualization technologies that could you could have the most inclusive local kubernetes experience you could have, and also we have support for different operating systems and different cpu architectures arm64 is getting very popular recently and minicube is also the only only local kubernetes experience that allows you to set a runtime I'll talk about the difference after runtime and a driver in the next slide.

A

But majority of the local kubernetes out there that you know or are similar to mini cube. They only have container d run time. Unique allows you to set the runtime, it could be cryo, docker or containerly mini cube. Also supports cni's, you can you can specify cni, you want. uh I don't know if you guys saw the the security talks on psyllium. Liz rice had a very amazing uh token and she was using psyllium.

A

uh I was very tempted to re-replicate her demo using mini cube, but I make it possible because, meaning you, you could just easily specify your your cni uh by the way the green ones are, the ones that we test them and uh green ones are the ones that do not have an integration test for them.

A

What is a driver in minicube drivers? Are this column?

A

Basically, if you want to install kubernetes, you need a linux and you need a linux box, so somebody has to give you a linux box, either a docker container or a vm, or basically a remote machine like ssh or a bare metal.

A

So that's the driver that you call that a driver minicube, whoever provides you that linux box, if you can install kubernetes on it, so you can in mini cube, start the driver with whatever driver you want and if you don't specify it, minicube will automatically select the best driver for you, based on your system, health and what you have installed and what is the container on time.

A

So, in this picture container entirely, these three ones container runtimes, is the engine that kubernetes itself uses to manage the life cycle of the containers inside a pod. So this is the container run time inside mini cube. It has nothing to do with the driver. This. This two gets confused. A lot by the people who use mini cube- and I want to just in this talk- make make it clear for everybody. So we never have this confusion again.

A

So the driver is the outer linux box that uh you, you per you get the linux box and container run. Time is the runtime that kubernetes uses to create pods for you, so that the pink one is your driver, and then there will be many cubes on xbox is the driver, and then inside of that will be container on time that could be cryo continuity or docker.

A

The confusing part is that driver could also be darker, so you could have it docker with a darker runtime or you could have a docker or a container due on time. So that's mostly confusing part, but I hope this slide uh clears the confusions once forever, but here's even more confusing things from a docker docker has so many news in the uh in the past few months or past year and kubernetes has so many news on docker past year.

A

uh In recently, in december 8 2020 kubernetes says we are deprecating docker and people like freaked out. It's like what does that mean, and so this is about the run time, it's not about the driver, so you could still have the docker driver with the container on time.

A

So, but what does that really mean? Does that mean you can never can no longer have a docker runtime anymore? No, it means the kubernetes no longer wants to maintain the code that allowed translating docker shim to a cry interface, cri interface.

A

However, it is possible to still have container docker, runtime and minitube chose to continue supporting docker runtime, not the driver. Why this graph shows mini cubes. Many use main users are local, kubernetes developers, the people who build an image and they deploy it to their cluster, and we have different ways to build an image. For example, in kind or k3d, there is a command called kind image load or k3d image. Load mini cable has eight different ways to build image. One of them is called docker n and you could see the mini cube.

A

Stalker amps is the fastest way to build images on the cluster. The blue one is the mini, cube, stalker end and the orange one is micro, kate and green. One is k3d and the yellow one is kind and the the lower the better the number so mini cube has committed to continue supporting the docker runtime, because it's significantly faster for the local developers and that's when you build an image.

A

You change it iteration on it, and you want to build another time and you want to save time on that, so mini cube is five years old. I was in the conference. I was talking to people, I I heard a lot of misconceptions and I hear them in misperceptions, and I hear them before in the surveys that we we collect from the users, and I want to go over the top mis perceptions in about mini cube.

A

It is true that three years ago, mini cube was very slow. In fact it was so slow. It was taking about three minutes to get started. uh We changed that two years ago we introduced the new drivers, we refactored a lot. We we introduced preload. I gave a talk on that in the previous kubecon, how we made miniqueue faster, but people still think minicube is slower because when they compare minicube with others, they compare the vm drivers with let's say k3 these docker driver.

A

So if you want to compare these two, you want to compare minicube's docker driver with let's decay, 3d use docker and another misconception is minicube by default: waits for the kubernetes cluster to be healthy.

A

If you say minicube start by default, we wait for api server to be up and running at cd to be up and running, and the coordinates to be answering the dns.

A

If you don't want to wait for that, you can say mini cube with false and it will be exiting as fast as others, but these are all like not very good ways to claim if munich will be slow or fast.

A

So we built a tool called time to case and what this tool does is measures how a local kubernetes tool, regardless with what it be uh performs, and we collect different metrics like when the tool itself claims, I'm done. Let's say minicube starts or current cluster create or k3d cluster create when they claim they are done.

A

That's the first metric we measure and then, when the kubernetes api server is answering and then we deploy an app and we measure when the app is actually running on it and then when the dns is answering and also you measure the cpu uh past two years, mini cube went from two minutes and 35 seconds to 21 seconds and here's a chart that we generate automatically in github actions per release. So this you could see mini cube against uh kind and k3d, so we're significantly faster than kind.

A

We are not as fast as k3d k3d is beating us and k3 is, has an advantage they're not using the scd they're using the sqlite for for database and which is a kind of you could say, a fork of kubernetes because kubernetes the real kubernetes uses at cd. It doesn't use sqlite, so that significantly makes it faster.

A

So this is the data and the misconception that minitube is law by data. You could prove it that it's it's a misperception.

A

The second misconception uh misperception is high cpu usage. It's true miku used to be extremely intensive on cpu three years ago and we significantly cut down on the and the cpu usage. uh Actually, you can check out the talk by priyawadua, improving performance of your kubernetes cluster in kubecon 2020, and she goes over how we did that in a lengthy talk and we measured the cpu usage again.

A

So the first one is your os idle and the second one is mini cube, hyperkit driver and third, one is making your virtual box and you can see mini cube, has lowest cpu usage against kind and create k3d and docker desktop.

A

So this is another misperception that it was true three years ago about mini cube, but no longer another misconception is like mini cube, is not a real kubernetes cluster. I see this a lot in the surveys. uh Services I wish uh making was a real kubernetes cluster, but it really is we're actually passing kate's conformance test. I have put the link to that. We pass a production case conformance test. We do not.

A

We do not recommend using minicuban production, but we do pass all the conformance tests and also you could actually configure every component of kubernetes in minicube. There is a flag called extra config, so you could pass basically any argument to cubelet cubeadm.

A

Basically any kubernetes component out there, you could configure it using mini cube, start extra config, so you could actually do any sort of configuration. Real world configuration advanced configuration in your kubernetes, so don't think of mini cube as a toy. It's actually a real kubernetes inside a linux box.

A

Another thing I hear a lot like I wish minicube had multi-note. We delivered this feature two years ago and not people. Many people know about it, but you could actually start making with multi-nodes very easily. You can just say: mini gibbs start dash, n2 and multi-node, or if you already have a cluster, you just want to add another note you can just say mini cube note at uh that would be another way for that.

A

Another one. That.

A

Is out there is like mini cube, cannot be multi-clustered, uh and that is not true as well. You can create a separate cluster with dash p. That's a little bit unfortunate that we chose the name profile five years ago for a cluster, that that is the source of the confusion.

A

So if you want to create a second minicube cluster, we call it unfortunately p, but we have stuck with that uh because of backward compatibility, uh we could just say immediate start dash, p p, one unique star dash vp two and you could actually configure them differently. You could say the second cluster I wanted with container d and the third cluster I wanted with cryo or different drivers and different, even different kubernetes version.

A

You could actually say start mini cube with a different kubernetes version. I actually missed to mention one of the mis perceptions, and that is people think you have to install the latest version of mini cube for the latest kubernetes version, or if you want an older kubernetes version, they think they should go. Install older miniki version, but that's not true. Also you, you could just do mini cube, start dash. Kubernetes version and pick a very old version to emulate what you have in production.

A

Another misconception is not good for ci, so that motivated me to create a new org called minitube ci, uh and this actually is a real world example with uh a get ups model that has example of mini cube in all of these clouds. In pro google cloud build github action, azure pipeline, travis, ci circle, ci and gitlab. Minicube is really well suited to be used in the ci, especially with the non-driver, which does not need any virtualization at all.

A

It's just straight on the on the ci machine that you have so, but if you want to see an example of using minicube in any of the ci, you could check out this report now I want to go over 20 pro tips for a mini cube. uh I will go over them very quickly and I apologize for speaking fast, and I know it's the last talk of the day, so um you're probably running out of caffeine, everybody, including myself, but I'll- go over this 20 tips that I've gathered that I have not actually seen.

A

Many users use that, but it's possible, you could copy a file into mini cube using minicube. Cp command is pretty neat and actually you can copy between mini cube nodes like the second one is an example of multi-node you're copying a file called atxt to the second node of a multi-node cluster, and you say which directory you want to copy that file up to for more information. You can check out the documentation, uh there's another way to copy files to minicube, and this one is very hidden.

A

You don't talk about it that much and the mini cube home folder is in your home, folder dot mini cube. There is a folder called files. Anything you put. There will actually be copied to mini cube on and every minute you start. This one requires a mini cube start, but uh that's why you have to run minikip start one more time to get this affected.

A

Another thing I have seen people not use this feature at all is like uh cni they're, not even aware of it, and you could uh try, let's say mini hubert psyllium, cni using makeup start or, if there's a cna out there, that we do not support it. Yet we support five six of them.

A

You could pass the yaml file pass to the cni and until we'll just.

A

Install that cni and and we do the c9 in a specific order- that the core dns will be configured correctly.

A

This is another one, specifying a different container run time mini cube, start container, run tom cryo one caveat to this: if you want to use cryo, you want to use mini q, pod man n, instead of mini cube, docker n, because cryo goes well with podman, and if you want to build an image against cryo, you could use spotlight another one is mini, cube, add-on list.

A

We have tens of add-ons and then some of them are maintained by google. Some of them are by third parties, some of them by kubernetes. They can give them a try, but this is an interesting one. Let's say any of these add-ons. You wanna, you wanna, use a different image for it. Let's say you don't want to wait for a mini cube to install to bump the helm, helms image, you could just say mini sorry, you could say mini cube, add-ons, enable images and then provide the image that you want to override to that add-on.

A

This is a it's an interesting tip for myself.

A

There's another trick that you could use if minicube has flags called memory and cpu that you could specify the memory and cpu you could actually pass the word max and it will use the maximum cpu and maximum memory that you could possibly get out of your system by default. Minicubes tries to get a sane amount of memory and cpu based on what your system is. We auto select it the better system you have, they will get the better chunk of it up to a maximum there's another flag, minicube dash dash user.

A

uh I love this flag. uh We we are. We have started collecting, uh uh writing and audits json to minicube logs and if you actually run miniature blogs, you'll see added table and see what user ran, what command on minicube at what time and how long it took this dash user flag is amazing if you are sharing a cluster with different tools.

A

If you have a few tools that they all sharing the minute you cluster, you could see who started the cluster who stopped it, who deleted it, who upgraded the kubernetes version of it, who applied a different cni on it. So you can see the the flow of the work. So, if you are sharing a cluster, I strongly suggest adding the dash user to all of the commands that you use for mini cube, meaning keep start, stop anything really.

A

That would be helping with the auditing later we have a new trick for mini q blocks in old times when something was wrong in minicube. If you had talked to us in github issues, we would ask you to give us this file that file this vlog this log, this one this one. It was five six different things you have to paste for us or find for us no longer.

A

Is that the case you just if you're on this command mini q blog file and what whatever file name you want, it will collect all the different logs from different places from different inside and outside mini cube and they'll. Put it all in one file. It's an interesting way to like create a github issue.

A

um There's another pro tip that I want you guys to try it. If you want to be a pro minicube user, there's eight ways of building images with mini cube and it's on our website, so give it a try.

A

There is another tool that I think many people do not know about. It is called mini, cube, cube, ctl command. So if you have not installed cube ctl, you could actually use mini cube, skip ctl. If you say mini cube, qctl dash dash get pods. You do not have install cube ctl, but this is another trick that I was amazed by the contributor who contributed to mini q. I was like this is pretty cool if you actually rename the mini cube's binary and call it cubectl, the minicube binary will no longer act like miniature binary.

A

It will act like ctl. So this is an interesting thing, because you do not need to download both mini cube and cube. Ctl anymore, you can just download mini cube, rename it called cube ctl. It will be cube, ctl, um so another trick. Another pro tip.

A

Another pro feature, I'm curious. Anybody ever use mini cube, pass a show of hands nobody. Okay. This is a very cool feature that nobody has used, so you can actually pop when we were doing a lot of analysis of what taking mini, cube's cpu usage. Api server of kubernetes is a beast. It takes cpu uh and do you need kubernetes api server? If you're a local developer?

A

Do you you need it? Sometimes you need it. When you apply, uh you need to apply your yam file, but after that you don't need it. Your apps is running. If you pause the kubernetes api server, your apps will still be running in the pods, so pause mini cube whenever you, whenever you don't want it, it will pause only the kubernetes api server, your apps will be still running.

A

It will save a lot of battery if you're in a coffee shop, you don't have a charger and if you pause, but still your app will be running inside the query cluster, we even have something better called auto pass in the previous one.

A

If you pause it and you want to apply another app to your kubernetes cluster, you have to do mini cube on pause. What, if minicube, would do that? For you like each time you apply a new yaml file, it will unpause itself and each time that you're not using the cluster, it will pause itself. You have an add-on called autopass.

A

This is a dummy diagram that I draw really I'm not very proud of what I draw sorry, but this is how it works is, if you have a transparent, reverse proxy, uh that takes care of uh auto pausing mini cube. But if you want to give this a try, mini cube add-ons enable auto pass a pro tip.

A

Another protip. We have a new driver that I'm pretty sure nobody used this ssh driver anybody used, ssx driver, okay, half of a hand went up, but it went down immediately okay.

A

So this is a way you can install kubernetes on a remote ssh or if you have access to a remote machine, you could say install mini cube on that remote machine. For me, all you have to do is just give us. The ip you'll go ssh into that.

A

Of course, you should have ssh access to it and it will bootstrap a kubernetes inside there for you another pro tip you can you can ssh to the mini cube cluster itself, using mini cube, ssh command, and if you want to only run one command and be done with it not go inside mini cube, ssh dash dash whatever the command, it is, let's say pwd.

A

This is another one.

A

uh You would be surprised that this is one of the features that people asked us to implement, and this is called schedule stop.

A

What does it do if you say mini cube, stop schedule, five minutes mini cube, stuff itself. In five minutes, white people asked for this.

A

Well, people use makeup in embedded uses like in ides, so some people, I say, close their ide, but they don't want to wait for it for skillet and that don't give the ide to stop mini cube. It's enough time for a stock mini cube. So this basically says if I don't hear anything from you I'll, stop myself in five minutes.

A

So in that use case, the ide was every three minutes was scheduling. Another mini cube. Stop for five minutes, so it's about basically refreshing it. So this is a way to uh schedule many cubes or stuff.

A

Another uh new feature of mini cube that people probably don't know about it is output json, so you can actually make mini cube talk to you in json instead of in emojis and text, and this is also used for for embedded use that the the format for json is actually the standard, cncf cloud event formats, the mount command.

A

If you want to be a pro user, try the mount command. What is monta command good for mouse command is good for if you want to have persistent data outside mini cube. So if you want to delete your mini cube cluster, but you want to keep the pods data somewhere on your own machine, you can mount a folder from your mac or your windows into the pod, and then you will have real persistent data.

A

I this is a pretty nifty feature, but I have not seen people use it that much because I don't see that sometimes I see github issues on them, but corporate search. This is an important one, because I get a lot a lot. So a lot of us work for corporations and corporations have their certs right. uh Who here has corporate service in their company yeah? A lot of us yeah and corporate stores are very annoying because you want to call google, it does not cover google because I need the search.

A

So how do you add your corporate search to mini cube search so basically copy it to this folder and many people take care of it. Basically, we want to copy it to this folder and then run when you keep start one more time. So I mean it copies. Those search into the chain of the search and kubernetes can talk to the outside world. Using your corporate search, another tip, you have two more tips left and I am running out of time. um Gcp auth is a new add-on that we have.

A

uh You know in old times when you had to. Let's say your your laptop is authenticated to gcp and your pods wants to be authenticated to gcp. You have to create this secret apply the secrets. Oh, it was such a pain, so we developed this add-on called gcp auth.

A

That automatically gets the credentials from your laptop and puts it to every pod inside your cluster that you will so they will automatically have access to gcp, just like your own laptop has and there's another pro tip. If you do not want a specific pod to get that secret, just apply this label to it. Gcps skip secret and you will not get the secret and there's a link for more information on it.

A

Another pro tip we have two more left delete on failure, so mini cube philosophy from the beginning has been be backward compatible and do not delete people's data. So if you start mini cube and we crash halfway a lot of the times deleting mini cube and starting it again- just works.

A

uh Sometimes it's just a virtual stitch in hyper. We get stock, something and something gets stucky just deleting it works. If you want me to allow me youtube to delete it on itself and try again, you can just say mini cube, delete on failure, and then you can delete itself and try again on a failure, but by default will not do that, because you want to respect your your cluster if you don't want to delete it, if you, maybe you have something interesting in there.

A

Try our load balancer is another pro tip. Minicube, I think, is one of the only.

A

Local kubernetes that has direct ip access to the load balancer in the vm drivers, not in docker drivers, because docker has its own limitation and very many more. We have three interesting links: minicube tutorials, minicube handbook and minicube faq check them out now back to docker. Docker has had so much news past year.

A

Okay, the recent news is: they are going to be a page tool for non-linux operating systems, and but they will be free and open for the linux and open source users and there's some blog posts on that uh in read this is I took a screenshot of the mini cube surveys uh before I come to the kubecon, uh so it's like why you use mini cube.

A

It's like docker, that's not going away looking for docker desktop alternative, because I didn't want to be tied to docker desktop bundle, especially they were forcing updates uh any oss alternative for docker desktop as a replacement for docker desktop. This is like people are flying us with messages on. Can you be a replacement for docker? Desktop, uh I mean we are: are a local development kubernetes tool? We are not meant for that, but we can be so. This is how there are two ways to use mini cube as a docker desktop replacement.

A

uh One is using the docker end method. What does that mean?

A

You can install docker cli, you could using brew or whatever, but do not install the docker desktop or quit docker desktop or exit it and then use the mini cube docker and if you, the mini cube, docker and your terminal will be pointing to the docker inside mini cube and mini cube is a linux box, which means it's always going to be free as darker claims.

A

So and then your docker cli, if this do docker, ps or docker belt, is actually talking to the mini cube stalker, it's no longer talking to the docker desktop or there's no docker desktop in this case, so you could easily actually replace docker desktop with mini cube. There's one method, one. This is the second method we introduced a new command in mini cube, called mini cube image, build a communicable image, which has many many sub commands like image, lit belt image load image delete image list.

A

There are many sub commands for a mini cube image command, but you could basically build your images using mini cube without docker at all. You do not even need to have docker cli installed, you could just say mini cube image, build dash, t and then pass to your docker file, basically the same syntax with docker build and underneath we are using build kit uh and until your image will be inside windy.

A

Now this is like a more uh mini cube native method. The other one is like you're, pointing to the mini the docker inside minicube, which either way it works. So there are two methods. A lot of people asked and I hope in this conference. I answered all of these users questions with these two methods and if I didn't answer it quickly, hit me up on slack I'll, make a better a more clear instruction.

A

There are some lessons I've learned in in mini cube in the past three years of being a maintainer, and we have worked with hundreds of contributors. We have had 675 contributors, and many of them are very active.

A

Some of uh some. There was a time that I was looking at the metrics. There are more new contributors to mini cube than kubernetes and they're, more active developers to mini cube than kubernetes itself, and this was based on the culture that people before me built it, and I tried to preserve it like thomas stromberg did a great job in in doing this, and I followed his lead like being inclusive in action like mini cube, tries to be as inclusive as possible.

A

We try to have support different languages and we try to have our github issues welcoming, so people can talk to us and we have we're in a very unique place that a lot of people who try kubernetes they come to meet you for the first time. Some of them are not very technical, but they are very interested to be technical. So it's like a very interesting energy.

A

uh Another thing that I learned in this past few years was recognizing the contributors and we developed actually tools to recognize the people who do any sort of contribution, even try, helping in github issues responding that and then I'll show you an example of that and another one is like scheduling short one-on-one meetings with the contributors helped a lot in energizing this. This amazing uh kubernetes, and also inviting to the mini cube exclusive meetings. We have too many exclusive meetings. It's inviting you guys to that has been a big part of our community management.

A

Here is one of the tools that we developed called pull sheets. uh Basically, each time we release mini cube in the release, notes we put a link to the charts and and who did what sort of help in what category in mini cube. So even if you're, not a technical person, but you want answered people's questions, you get your name uh recognized in this leaderboard. It's going to be an amenity website forever, so we try to recognize contributions.

A

Another maintainer lessons that I have learned is embracing the human part of the softwares.

A

This this slide was very important for me, and I wish my talk was title differences. I could have dedicated the whole talk to this, I'm so passionate about this one, but I could only fit it in one slide and go over it very quickly.

A

What I've learned that some of the stereotypes about software engineers might be not too wrong. There's there's some our profession attracts different uh type of personality and very neurodiverse personalities, people with ocpd adhd cluster c and working with these people. Learning them on unleashes the power of of maintaining a big project.

A

If you learn that some people get obsessed on something and they cannot let it go, that person could be amazing if you want order in 1.5 million lines of code, but you also want to learn what is the strong part of this person? What is the weakness of this person or another type of person?

A

He might miss out paying attention to the biggest thing, but has the most creative mind, and you also need that type of person or you have the cluster c type of person who worries all the time if you want to implement anything worries all of time about the security implications of this, and the art of this is how we balance all of this. What is the budgets that we have? What is the budget for security? What is the budget for creativity and what is the budget for order it takes?

A

um I wish I could have more time to talk about this. There's other talks in kubecon that they talked. I love to touch this topic a lot, but I really learned a lot in the past three years that learning the humans around. You goes a long way to to to maintain a healthy project, and once you learn your humans once you learn your contributors, you know what type of health they might need. Some need assurance some need, keeping on track. Some need getting started.

A

Another lesson that I learned is don't take anyone for granted. Some people come to this open source project to take a break from their day jobs. The worst thing you could do to them is like take them for granted that you're yeah you're supposed to do this. Like you, that's your thing uh not taking anyone for granted also is a lesson that I've learned and also another uh lesson is like do not pronounce. This is another lesson that uh I've seen a lot of people in open source.

A

They come work really hard for two years and they burn out. There is a actually good talk in this stock. This kubecon, I schedule I I I recommend watching it on youtube uh check out the vulnerable tale of pronouns by julia simon this. This is important lesson that I learned in the past three years of being in mini cube.

A

uh Another lesson that, as a minicube maintainer, I learned was like learning when to say no, and that was very essential for this old big project to keep going. One is the spam contributions. I don't know if you guys have seen any spam contributions. People literally write bots to make contributions to projects.

A

Another thing to say no to is a promotion driven development. There is a amazing twitter discussion, promotion, driven development, people, write codes just to get promoted and you could just smell it and and if it ruins a project's future, it's- um and I have learned to say no to that another thing to say no to that. It's hard to maintain god, because we're like harder to maintain the code does amazing things but makes the project so complex that only that person could uh maintain. That, of course, do not emerge or accept any contribution.

A

You do not fully understand, there's some special interest people. I have seen that people with some special interest come to your project. They want to try to get this into their project. It's another place to say no another place uh to say no is potential security liability.

A

I have seen a few times. People wanted me to add a future that I saw potential security vulnerability like that. This is a third-party image in this code who can guarantee the next five years. This image is gonna, have the same integrity as, as today like this, this project has been five years. Old people have trust in it, so this is another place to stay hard. No, even though it adds a feature and also another place to has a hard, no is when somebody has a new feature without integration test hard.

A

No, this is amazing lessons I've learned by mistakes or by from others.

A

Another things is that I've learned is guarding against your own bias.

A

Unconscious bias is unconscious, it means you, don't know you have it and there's like at least seven of them that I strongly recommend take a wikipedia page and read about it and then and put that in your mindset.

A

Another main third lesson is like github issues is an amazing source of information for future work and improvements. When I look at the github issues, I see that the place that I learn from the users and when I am doing pr reviews is where the users learn from me. So they're two-way communications, but github issue is where I am supposed to learn from the users and for that reason we actually developed a tool to try edge the github issues.

A

It's called triage party, it's a multiplayer game that actually a few other cncf projects are using it uh and we started at minicube. The other projects are doing it and basically this helps you to try as the issues more effectively uh using a crowdsourcing technique and there's a different categories of of triaging. For example.

A

Let's say somebody uh creates an issue in mini cube and say: oh there's this problem with this, and I asked them for more information and then two weeks later, they answer with more information, but that gets lost in 500 issues and mini repo.

A

Wouldn't it be nice that a tool would tell me that the person who you asked for more information now has more information and they'll be in a different category. That's one thing that we we do. We did in miniqueue many of another interesting that we did is we translated minicube to multiple languages and youtube. Now speaks your language, uh so you can wanna. If you wanna, add your own language to minicube. It's super easy. It's a json file just translate the json to your own language.

A

Currently we have french, german, spanish, chinese, french and japanese and korean and polish.

A

Another thing I like to touch in this talk is the mini cube integration test. There's a challenge in minicube integration test and the challenge it is. We support many many platforms and many many runtimes and many many drivers, and we also have legacy, features that we do not want to break five years of user base, features that if we break them, we break somebody.

A

So those are the challenges and we have different types of testing unit, testing, integration, testing, kubernetes performance, test, pressure, test and translation test.

A

Currently, I do not know any other project in this cncf world that has as diverse testing environment as mini cube. We get tested in gcp, azure aws for arms, 64., equinix, bare metal, max stadium, which are basically physical, mac machines and github actions, and also we have alibaba in progress and that's I would say this is the most powerful part of mini cube that we get tested in many many uh locations and in many different drivers and runtime combinations.

A

Another big challenge for minicube was flaky tests. We have many many tests, hundreds of tests and on the pr you will see a failure and the failure would say this test is failing. So how would I know if this is a real failure? That's causing is caused from this pr or uh how do I know if this is from uh afflicted, so we've actually built a tool that uh that measures the flakiness of a test and comments on the pr and tells you the the flakiness, and that's also in our website.

A

You could check it out the the charts for mini cube flake base, and this is based on a tool called gopog, go poke. We built a tool for goliath that converts the integration test to html, but this has helped us to make the pr reviews much faster. You can see a a test is a flake or not very quickly.

A

Another thing that we have done in minicube. We have created a mini cubot. That makes a lot of the prs for us that we used to do by ourselves like bumping the golang bunk, bringing the kubernetes version bumping this bumping that all of that is like done by a robot. Now uh we don't have too much time. I think we actually ran out of time. uh I believe- and I I'm very amazed that you guys stayed because I didn't pay attention that be around that time.

A

uh But if you want you can follow me cube's twitter, we have a mini cube, bot called mini, cube dev. It automatically pushes updates to twitter and whenever there's a new mini cube release so follow many cubots and for future we're going to have vm driver for arm 64.. We have arm64 support for docker driver, but but not for vm driver yet, and we also have our plans for maybe unifying the the iso and the container-based damage with that. Thank you very much for staying.

A

uh And I'll stay for, if you have questions I'll be more than happy and sorry for running out of uh way over time.