Conf42 Cloud Native 2021 - Online

Autonomous Multi-Cloud serverless deployment and optimized management

Video size:

Abstract

The dynamic development of Cloud Computing with the introduction of novel Cloud computing models like serverless creates new challenges for Cloud deployment.

This presentation describes how to implement Multi-Cloud native strategies using advanced an open source framework that allows for Cloud-agnostic Multi-Cloud deployment and optimized management of the serverless applications based on flexible monitoring, context aware maximization of the application owner’s utility of the deployed serverless components, and autonomic reconfiguration based on the application’s current execution context.

All stages of the cloud deployment planning and designing process will be shown. Also, the key execution steps will be provided.

Summary

  • The melodic platform is the single universal platform for the automatic multi cloud deployed. It can be deployed to the Azure AWS, Google Cloud, to any OpenStack cloud providers. Everything is done automatically and even more the usage of the resources is optimized. You can consider melodic as your smart and autonomic DevOps operator.
  • How to automatically deploy own application by melodic platform. I will perform deployment of spark based application. We will monitor application metrics and observe reconfiguration process. Here is also possible to run application in simulation mode.
  • Genome is a big data application which performs some calculations and safe results in AWS as free bucket. Genome's performance is managed by Spark. Melodic creates proper number of spark workers as virtual machines considered our requirements. Thanks to measurements of application metrics, melodic makes a decision about creating additional instances with workers or about deleting unnecessary ones.
  • Melodic creates proper instances, remains them or deletes. If you want to have more detailed view, it is possible to see the process view using Kamunda. Grafana is tool for monitoring displaying statistics and metrics. We can use them for monitoring performance of applications deployed by melodic.
  • Melodic is fully open source so you can download the melodic here. It is released under the Mozilla Public License 20 so it can be used and challenges. We are also looking for the volunteers developers, open source developers. If you want to work on the interesting process then please join us.

Transcript

This transcript was autogenerated. To make changes, submit a PR.
Hello, my name is Pavusk and together with my friends Alaranievich, we want to show you the most in innovative multicloud deployment platform. Yeah, I know I'm out of that platform or one of the coauthors, but I really really think that it's a very innovative things. Let's start. What is the melodic platform? It is the single universal platform for the automatic multi cloud deployed. So you can just simply automatically deploy to the Azure AWS, Google Cloud, to any OpenStack cloud providers and also to one steps cloud provider and other cloud provided everything is done automatically and even more the usage of the resources is optimized. So our platform is selecting the most optimal set of the resources, the most optimal cloud providers to deploy your application. How it works the melodic is probably the simplest and the easiest way to use the multi cloud. You don't need to learn Azure AWS, Google Cloud platform OpenStack, but you have the one cloud agnostic independent models of the application and it can be deployed automatically to any cloud provider. It supports virtual machines containers serverless function which I especially covered on today's presentation and also big data frameworks and we can deploy that to the different cloud providers. As I already said, the deployment is fully automatic. So it's probably the only one multicloud platform which supports fully automatic deployment and for sure the only one platform which is doing advanced optimization of the cloud resources using the machine learning based algorithm. So first step is to model the application and for that we have deployed camel cloud application modeling and execution language which allows to model both application and infrastructure. And even more it allows to model the requirements, constration and optimization goals for the application. So everything is modeled in one model and this model is used CTO deployed application to various cloud provided and for the optimization proposals. It is quite similar language based on Tosca, but it extends Tosca a lot, allows for the modeling components, connections between the components, security rules and so on. And of course to model the infrastructure requirements so we can model the requirements, constraints and the value of the utility function which is based as a goal for the optimization. The utility function set what is the best deployment. So how we know that the given deployment for the application, given set of the resources is the best. One of the criterion is of course the cost. But if we have only the cost as a goal or as utility function, then the best solution is just simply to not deploy anything because in that case the cost will be zero. But in reality we want to deployed our application. So it is usually the trade off. But before this optimization. We need to know what is the context of the application, so how many resources are used and what is the performance, what is the cost and so on. And we are doing that by the metric collection. Then we need CTO have the utility function. So we need to know what is this past deployed. And in melodic it is possible to create in a very flexible way the utility function. And this utility function is based on the business value. So not on the cpu usage, memory usage, but on the average response to the user, average workload and so on. It is possible to add any custom metric for the application. And this custom metric measured by the custom sensor can provide the business related metric to the platform which is used for the optimization. And as I said, it is usually a trade off between the cost and performance, cost and availability, cost and other requirements. And everything is modeled in the camel model and automatically optimized by melodic. So you can consider melodic as your smart and autonomic DevOps operator. How it works first step is manual, so you need to model application in the camel model. But the good news is that it is needed to be done only once. So if you have modeled your application, set initial values of the parameters, then everything is done automatically. So after deploying the model, melodic is calculating the initial deployment plan for the application. Deployment plan, I mean which cloud providers to use and how many resources, what type of the resources and so on and so on. Everything is calculated automatically. After finding the most optimal deployment, the application and infrastructure is automatically deployed to the selected cloud providers. So virtual machines and other resources are provisioned. Then the components are deployed fully, automatically. And then the metric collection and monitoring is started. So the melodic platform is collecting the metrics and based on the values of the metrics, melodic makes a decision if the new deployed new solution should be reasoning to start rezoning process to find the new solution. Usually the decision about starting new reasoning process is based on the threshold. So if the given metric has the value above the certain threshold, then the new reasoning part is started and melodic is looking for the new solution. And if the new solution is found, then it is automatically deployed, but it's not deployed from the scratch. But the current one solution is reconfiguration. So new resources are added or removed if necessary or replaced. But everything is done on the fly. It's fully automatic process and how melodic is built. I think it's very interesting, especially for the developers. For the cloud native solution, the melodic has the pure microservice architecture as you can see. And these microservices are orchestrated by the PM process. We have implemented all of the logic as a business process using the kamunda and for managing the components and invoking the components. We are using Mu ESB as ESB and activeMQ with the esper as the monitoring plan. Thanks CTO that we are very flexible. We can change the logic just simply by drag and drop in the BPM tool. We avoid to hard code too much logic into the system. We have evaluated different integration method point to point so directly from the BPM to the certain microservice cube based integration ESB. But finally we have decided to use the ESB with the BPM integration as it is implemented currently in melodic. Yeah, we have an extended comparison of the different methods and how they fulfill the requirements including some kind of the points calculation and preparing the planning. And as I said finally we have evaluated free USB solution and four BPM engines. But the final solution is built based on the mulesB and Kamonda BPM workflow. We have documentation why we have selected that tools and it could be also very interesting thing to take a look if someone wants to build this type of the solutions. We have compared both ESB and BPM solution and based on that we have built the whole platform which is then used to deploy the applications. And today I want to show two examples of the use case application and then Alicia will go through the melodic platform and show the r1 deployment. One application is image recognition on the fly. The images are provided by the special augmented reality glasses and directly connected to the cloud and we are using serverless function which classify images almost in the near real time and melodic decides how much resources or how many instances should be run in the given moments. And also the second planning part is deployed using the spark big data framework which is also managed by melodic. Melodic decides how many workers should be used. The second use case is big data application spark application. It is the application used by the one of the polish universities to analyze genome data. It compares the genome of the given person, CTo the reference genome data to find the mutation and the genome data are collected from the persons and then the processing is started after sequencing to the digital form of the genome and the operator is starting the processing based on the number of the patient and data to process starting this processing, melodic is looking for the parameters of the training, so collected the metrics and checking how long it takes to process the data and if it process too long the new resources are added and the process repeats. So melodic continuously collect the metrics and checking what is the expected estimated time for the finishing job and if the time is too long, then again the new resources are added and if the time is below the expected time. So the business goal the reconfiguration is keep stable and the data are process and at the end the data are removed as a summary how the workflow in the melodic looks like. So we model our application in camel, then deploy the melodic platform, then submit the model to the melodic platform, press big green button in the melodic, Allah will show you how to do that and then melodic automatically find the best solution and deploy your application. And after that just connect to your application and enjoy. And now allah will present the demo of the melodic. So the demo of the platform and also the deployment of the application, how it is done and how to handle that. So thank you and allah please go on. Now I would like to present you how to automatically deploy own application by melodic platform. I will perform deployment of spark based application. We will monitor application metrics and observe reconfiguration process which is done by melodic for reasons of optimization. My melodic platform is installed on virtual machine on AWS and it is up and running. I'm logged in. Melodic users are managed by LDAP. We have three possible roles of users common user he can perform application deployed admin user. He manages of users accounts and also has all privileges from common user and technical user. He is used only internally by melodic components and he is not important from client's point of view. The first step in melodic usage is the defining of cloud settings. In provider settings menu part, we can check and update provided credentials and options as we can see in cloud definition for provided view, filling these values is required in order to perform successful deployment because they are based in contact with providers. For example, by creating virtual instances on my environment, I have already defined these values for Amazon web service and for OpenStack providers. In these definitions we provide cloud credentials and properties, for example settings for Amazon Security group or set of private images which we would like to use in our deployments. When our platform is prepared configured, we can go to deployment bookmark. Today I would like to deploy genome application which was described by Pavel a moment ago. Before deployed we need to model our application with its requirements in camel model which is human, understandable and editable form. After that such model is transformed CTO XMI format form understandable formaltic we upload this file here by drag and drop. Now our models is being validated and after that it will be saved in database. In a minute I will be asked for fill values of AWS developers credentials. Providing these credentials is required in order to save results of our genome application in AWS as freebacket, but in view of security reasons we shouldn't put them directly in camel model file. So we use placeholders in camel file and after that we need CTO provide these values here. In this case it is not the first upload of such model on this virtual machine. So these variables already exist in dedicated secure store. I can verify them update if they were changed and after that choose save button. In the last step I need to choose which application I want to deploy and which cloud providers I want to use. Here is also possible to run application in simulation mode. Simulation mode is the case when we don't want to deploy real virtual machines on provided platform but only check which solution will be chosen by melodic. We manually set values of metrics in simulation part and observe the result. But today our aim is to perform real deployment of genome application so I leave this option turned off. We would like to deploy genome only on AWS so we chose this cloud definition. Thanks to that melodic has credentials for this provider. After that we can go to the last step here where starting deployed is available. After starting the process, in a minute we are moved cto the deployment process view. Here we can observe the progress of that. In the meantime, I would like to briefly describe application which is being deployed by melodic. Now Genome is a big data application which performs some calculations and safe results in AWS as free bucket so we need to provide developers credentials to AWS. Genome's performance is managed by Spark. In genome application we use Spark as platform for big data operations which are performed parallel on many machines and managed by one machine named Sparkmaster. Sparkmaster is available by default on melodic platform. Melodic creates proper number of spark workers as virtual machines considered our requirements from camel model thanks to measurements of application metrics, melodic makes a decision about creating additional instances with workers or about deleting unnecessary ones. Spark divides all calculations named tasks between available workers in order to optimized application performance and cost. Please let me come back to our process. Phishing offers is the first step of deployment process. We have information about current total number of offers from previously selected providers. So in this case from AWS from these owners. Verlodic will choose the best solution for worker component after choosing this box or offer option from Mani which is available here, we are directed to view of all currently available offers. There are cloud with my credentials and also with my properties for security group and for filters for our private images. Also we have here hardware with information about cores, ram and disk and available locations where our virtual machines could be located and the last element here images. There are only private images visible here but of course all public images are available for us. Now I come back to our process view and we can see that the next step of process is generating constraint problem. Constraint problem is generated based off our requirements defined in camel model. In simple process view there are visualized all variables from constraint problem with the domain values for genome worker, cardinality worker course and provider for spark worker. Detailed data are shown after click of this box and here are presented list of variables with additional information about component type, domain and type of this domain. Utility formula it is used for measure utility of each possible solutions and choose the best one list of constants with types and values. They are created from user requirements and are used in melody calculations. Here we can see for example minimum and maximum values for cardinality of spark worker instances or the same type of restriction for number of spark worker cores. So we can see that in our deployed we would like to have from one to maximum ten workers and the last element here list of metrics with data types and values initial values. They describe current performance of this application. Thanks to them, melodic can make a decision about triggering the reconfiguration process which means creating new additional instances or deleting not fully used ones. Thanks to metrics, melodic can do the most important task which is cost optimization. We back to process view. When constraint problem is generated, it is time for resoning. Melodic finds here the best the most profitable solution for the problem defined by us. When rezoning is completed, we can observe information about calculated solution utility value and values for each variables. In that case one as worker cardinality for worker cores and provided for spark worker from zero index. So it is AWS. The next step in process deployment is deployed here melodic performs operations based on calculated solution. This solution is deployed for each application component. Melodic creates proper instances, remains them or deletes. If you want to have more detailed view, it is possible to see the process view using Kamunda by choosing advanced view button from upper left corner. From this view, Kamunda is tool for monitoring and for modeling processes in BPM and standard and for management of them. I log in by the same credentials as for my melodic platform and in order to see detailed view in Kamunda I need to choose running process instances and after that process to monitoring from the list. And now we can see view of chosen process with all the variables and also detailed view of the whole process with each steps. This view is for more technical users. It could be useful for example during diagnostic of some problems. We can see that now we are even here. So it is the end of our process. In order to verify this fact I go to UI again and yes, we can see that our application successfully started. So the deployment process is finished and I can check the result in your application bookmark. In this view there are deployed list of created virtual machines and functions. Genome application requires only virtual machines. We can see that melodic creates one virtual machine. Thus far this machine is created in AWS EC two provider in Dublin. What is more we have here button for web SSH connection which is really useful in testing process. When I successfully deployed spark application by melodic, I can go to Grafana. Grafana is tool for monitoring displaying statistics and metrics. We can use them for monitoring performance of applications deployed by melodic. Each application has own metrics and own parameters to control. So we need to create dedicated Grafana dashboard for each of them. Also genome applications has own Grafana settings and we can see them here. For now metrics from our applications are not available yet. We can see only that we have now one instance so one worker. In the meantime we can control our application in Sparkmaster UI. Sparkmaster is built into melodic platform so we go to the same IP address and 81 81 port in order to check the spark master UI and here we can observe a list of available workers. After refreshing of this view, of course we can see that now we have one worker and one running applications and also one driver. So now all tasks are sent to this one worker by our spark master. It is situation after initial deployment decision about creating the new ones or deleting. Some of workers are made by melodic based on measured metrics. In such situation new process is triggered and it is named reconfiguration process. I think that now we can go again to our grafana dashboard and we can see that metrics are using correctly calculated and based to the melodic because they are visible also in our grafana view. Color of limitation of traffic lights inform us if application will finish on time. Now we can see the first estimation so it can be not correct. I think because we have no enough data for a good estimation. So we need to wait a minute and even now we can see that our light is red. Also we can see that our time left is not enough to finish our calculation on time. Because the initial indicated time is indicated in camel model. In this case it is equal to 60 minutes. We can observe how many minutes left from this time period. Under this time left value based on current performance. It is calculated the estimated time left on the left. On the first chart we can monitor number of instances. So now we have one node. So one worker. In the bottom ones it is presented number of remaining simulation. This value is decreasing with performing next task by Spark. On the right on chart named number of cores, we can see value of minimum course needed to finish calculations on time and current number of cores under total course value. The green one is the value of required number of course and the yellow one means current number of them. Now melodic claims that we need at least four cores and even now six cores, seven cores and we have only one. Also estimated time is higher than time left. We can see red light. So there are signals that our application needs more resources. In such situation, melodic makes a decision about triggering reconfiguration process. So we can suppose that in the background reconfiguration process should being done in order to verify it. I back to our melodic UI and I go to process view. And here we can see current process and it is our reconfiguration process. In reconfiguration process, melodic doesn't fetch new offers and uses the same constraint problem as for initial deployment. For these reasons the first step is rezoning. As result, we can see new calculated solution which will be deployed. Now melody claims that two workers will be needed and this solution is now deployed. So in a minute we will see our new worker. Oh yes, even now our new worker should be visible because the reconfiguration process is finished. I can verify this fact also in your application part. And yes, now we have two virtual machines, two workers and this is the new one from our reconfiguration process. Also I can check this fact in our Sparkmaster UI. I need to refresh this view. And now two workers are available. We have two life workers. So now we can see that Sparkmaster divides tasks between these two workers in case of genome. In the first part of performing calculations, additional workers are created. As far as melodic measures that effectiveness of application is too low. In the final part of performance of spark jobs, melodic makes decision about deleting unnecessary instances when it is feasible that applications will finish on time. And now we are in our initial part of the whole process because we have, as I mentioned, we have 60 minutes to perform the whole process. So we are at the beginning of them and now additional workers are being created because of effectiveness of our application. Now I go to Grafana view. We can see that now we have two workers, two nodes. Next tasks are done and we can see that now our estimated time is cloud to time left and even now melody claims that it will be possible CTO finish the whole process on time. But now our estimated time is bigger again. So we can suppose that in a minute our time will be red again. Probably we will see the new reconfiguration process and the whole process is using performed to the moment where our estimated time will be enough for us, enough for our requirements. Thanks to that, finishing the whole process in our expected time will be possible, right? So we successfully observed the configuration process of spark application. This is the end of spark application deployed done by melodic demonstration and we can see that the whole optimization process is done fully automatically. Okay so thank you very much for your attention and this is all from my side. Okay thank you AlA for the presentation. Just a few words from my side about the melodic. Melodic is fully open source so you can download the melodic here. The source code is hosted on the OW CTO GitHub so you can download the code. It is released under the Mozilla Public License 20 so it can be used and challenges. Anyhow, welcome. We are also looking for the volunteers developers, open source developers. So if you want to work on the interesting process then please join us. We are currently developing further melodic in the scope of the morphemic platform with new context like polymorphic adaptation and proactive adaptation. But I will tell more about that on the next session. Probably something in the future. One more thing, please take a look on our website and also visit our Twitter, LinkedIn, Facebook and follow us on the social media. Thank you very much. Thank you for the invitation to have the session on the cloud native. I hope you will find that interesting and welcome to the melodic and I really invite to join to our community. Thank you very much.
...

Paweł Skrzypek

CTO @ 7bulls.com

Paweł Skrzypek's LinkedIn account

Alicja Reniewicz

Full Stack Engineer @ 7bulls.com

Alicja Reniewicz's LinkedIn account



Awesome tech events for

Priority access to all content

Video hallway track

Community chat

Exclusive promotions and giveaways