Demystify Huawei Cloud
Topics that we will dive in are SaaS (Software as a Service), PaaS (Platform as a Service), and IaaS (Infrastructure as a Service) on Public Cloud platform and how it brings convenience to the users.
In this session, our Huawei Cloud expert – Praveen will bring everyone to venture into Huawei Cloud and back. Praveen will also guide everyone to look into a simple demonstration on how to purchase an Elastic Cloud Server (ECS) / Virtual Machine (VM) on the Huawei Cloud platform.
Without further hesitation, let’s join the session and gain some insight from his perspective!
Mr Praveen Nair
Trainer at Infosyte
Praveen started off his IT career as Project Engineer taking care of the network infrastructure of the organizations across Asia. His areas of expertise are storage, cloud and big data where he is tasked to conduct the respective training when he joined Infosyte in 2017. He is very informative and engaging with his students to keep the learning session interesting. As a result, Praveen became one of the favourite trainers for South Africa students where the students specifically requested for Praveen to be their trainers.
Huawei Cloud Biocloud is a public cloud platform similar to vendors like AWS, Microsoft Azure, and Holistics. It is a foreign vendor but operates in a similar way to other cloud computing services. Huawei Cloud is a vast virtualization environment offering various services including network virtualization, storage virtualization, and compute virtualization. This allows users to build their environment on the cloud, using Huawei Cloud as a virtualization engine. Customers can purchase services, rental services, or build their own virtualization environment on the cloud. While there are physical services available on Huawei Cloud, this post focuses on virtualization services. Pricing for virtualization services is affordable, even if some other services are more expensive. In summary, Huawei Cloud is a comprehensive virtualization engine for building and purchasing cloud services.
To discuss the core elements of Foyer Cloud, it’s essential to understand the three primary service models: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). IaaS provides a virtual infrastructure to build your services. By procuring the base infrastructure from vendors like Huawei, you can build your servers, databases, and applications in that environment without worrying about the actual hardware. Network engineers, server administrators, and storage administrators usually provide the base infrastructure for IaaS.
On the other hand, PaaS is where developers work. Software developers and application developers use PaaS as it provides them with a pre-built product. Developers do not need to concern themselves with the intricacies of the network environment, server workings, or other technical details. PaaS provides an efficient software installation to run or build applications.
Lastly, SaaS is for end-users or customers. When customers come to the cloud, they can find pre-built software on the cloud that they can purchase and use in their environment without worrying about building the software from the ground up or taking care of the network. The software is pre-built, and customers can use it as per their requirements.
Foyer Cloud provides all three primary service models. However, some cloud providers like AWS and Microsoft Azure offer only one or two of these services. Providing all three requires a lot of hardware, background, and infrastructure. Therefore, Foyer Cloud’s ability to provide all three services makes it a unique cloud provider in the market.
Let’s now turn our attention to the computing services offered by Hawaii Cloud Journey. Compute services are essentially designed to provide users with CPU and memory resources. At the top of the list of compute services is Elastic Cloud Server (ECS), which is essentially a virtual machine on the cloud. It possesses all the attributes of a physical server, but instead, has a virtual CPU and memory. Similarly, network components such as NIC cards are also virtualized. ECS offers a range of specifications, depending on the user’s specific requirements. Whether you are looking to migrate your entire environment to the cloud or develop an application, ECS has the necessary specifications to meet your needs.
Another compute service worth mentioning is Auto Scaling, which goes a step further by automatically scaling your resources based on certain policies that you configure. While this is a valuable service, it is important to configure the policies properly to avoid any potential issues. Auto Scaling can scale servers instantly, adding or removing them as necessary. Alternatively, it can also scale dynamically, based on CPU or memory utilization levels. This service is particularly useful for e-commerce environments, where server demand can spike during a sale.
A stateless environment is required for Auto Scaling to function optimally. Servers that store information locally (stateful environments) can cause issues. The last compute service we will discuss is Image Management Service (IMS), which can be likened to templates used in VMware environments. IMS allows users to create a template of a running server, which can then be used to deploy multiple servers equipped with the same applications. This service simplifies the deployment process, reducing workload and freeing up time for other tasks.
In summary, computing services are an essential component of cloud computing. While the above-mentioned services represent only a few of Hawaii Cloud Journey’s offerings, they serve as an excellent introduction to the range of services available. Other services, such as bare metal servers, are also available.
Network services are a fundamental component of cloud computing. Among these services, the Virtual Private Cloud (VPC) is a core service that allows you to create an isolated network environment to build your cloud architecture. The VPC provides a simple way to allocate IP addresses, subnets, and servers, which can be assigned manual IP addresses or through DHCP. However, it is important to note that simply using the default VPC or network configuration may not be sufficient for most use cases. Careful planning and consideration of the network environment are crucial to ensure optimal performance.
In addition to the VPC, Huawei Cloud provides other network services, such as NetGateway and Network Address Translation (NAT). NetGateway is a product that allows for source and destination network address translation. It enables your server in Huawei Cloud to access the internet if you are using a public NetGateway. Destination network address translation allows traffic from the internet to reach your services inside Huawei Cloud. These services are intricately linked to the public IP addressing scheme, which will be discussed in more detail later. Port mapping is another useful feature of NetGateway that can help you save IP addresses and reduce costs.
Elastic IP is another network service that is commonly used in cloud computing. Essentially, it is a public IP address that can be assigned to your services in Huawei Cloud. Elastic IP allows your servers to access the internet, and it is particularly useful for load balancing and other public-facing services. Huawei Cloud provides a range of public IP addresses that can be assigned to your services, along with a choice of bandwidth options. The cost of these services is based on the amount of bandwidth used, so it is important to choose an appropriate level of bandwidth based on your needs.
Overall, network services are a critical component of cloud computing, and careful planning and consideration are necessary to ensure that your network environment is optimized for your specific use case. Huawei Cloud provides a range of services that can help you build a robust and efficient network architecture, and it is important to understand the capabilities and limitations of each service before making a decision.
When it comes to cloud storage services, there are three main types that you’ll typically encounter: elastic volume service, scalable file service, and object storage service. Huawei offers all three of these storage services.
Elastic volume service is essentially block storage, which means that it functions like a hard disk that you attach to a server. After attaching it to an ECS, you’ll need to initialize the disk and format it before you can start using it. In essence, elastic volume service is a virtual hard disk.
Scalable file service, on the other hand, is similar to network-attached storage (NAS) in the cloud. This type of storage cannot be used alone, and you’ll need to mount it to a server in order to access it. You can connect multiple cloud servers or ECSs to a single scalable file service to enable multiple file sharing. There are two types of products within scalable file service: SFS and SFS Turbo. SFS can handle around 10,000 connections, while SFS Turbo can handle 5,000 connections at a faster performance speed in terms of IOPS and other metrics.
Object storage service is the type of storage that powers cloud-based storage systems like Google Drive, as well as video-on-demand platforms like Netflix and Disney+. It can be accessed using RESTful APIs at the HTTP level, without requiring a server to be mounted. However, you can still connect backup data to an object storage service and access it through a server environment.
While elastic volume service and scalable file service have been around for decades and are therefore quite familiar, object storage service is a newer storage option that has unique features and capabilities. It’s important to note that traditional applications may not work as well with object storage service, so it’s best suited for newer, cloud-based applications.
Overall, Huawei’s storage services are a great option to consider. Object storage service is the most affordable option, while elastic volume service is the most expensive. To learn more about these storage services, explore your options within Huawei’s cloud environment.
As we navigate Huawei Cloud, logging in is the first step to accessing its services. There is a registration and login button located on the side of the page. Upon registration, you will need to enter your credit card details as payment is required to use the cloud environment. Huawei Cloud offers free packages, allowing users to try out various services available on the platform. These services can be accessed through the console environment, which is where all the operations take place.
It is important to ensure that you are in the correct region before deploying your services. Huawei Cloud has multiple regions available, such as Hong Kong, Bangkok, Singapore, Johannesburg, and Mexico. Choosing a region closer to your location will ensure faster service deployment.
Huawei Cloud provides a wide range of services, including computing, storage, network, and management applications. The service list can be found in the left-hand corner of the console. To create your first server, we will focus on the Elastic Cloud Server (ECS) service. It can be found by typing “ECS” in the search bar or by selecting it directly from the list of services. It is also possible to add frequently used services to the favorites bar, allowing for quicker access.
The favorite bar can be accessed from any service within the console, providing convenient navigation throughout Huawei Cloud. Once in the compute service, we will use the Cloud Server Console to build our server. This console contains everything required to create and manage servers, such as disks, images, and auto-scaling services. We will be using the ECS service, which can be purchased by selecting the “Buy ECS” button located on the right-hand side of the console.
In summary, Huawei Cloud provides a comprehensive range of services accessible through the console environment. By registering and selecting the appropriate region, users can easily deploy services and manage their servers. The Elastic Cloud Server (ECS) service is an excellent starting point for those new to Huawei Cloud.
When configuring your system, your first consideration should be the billing mode. Typically, there are three options: yearly, monthly, or pay-per-use. Yearly and monthly billing options require prepayment for one month, with the cost remaining fixed regardless of server usage. Pay-per-use billing is an hourly rate that charges you 0.101 USD per hour for server usage. It is important to note that shutting down the server while using the pay-per-use billing option will result in a reduced charge, as CPU and memory usage will not be billed.
Another billing option available is the spot price, which is generally used for testing environments. While it is cheaper than the yearly or monthly billing options, it is subject to price fluctuations based on supply and demand. If there is an excess of supply, the price will decrease, while a high demand will result in an increase in price. If you use the spot price option and the price increases beyond what you paid, your ECS can be reclaimed and provided to someone else who purchased it for a higher price. As a result, your server will stop working. The spot block option provides hourly-based pricing for those who want to test a specific environment for a set amount of time.
When selecting a region for your service, it is important to choose one that will provide optimal reliability. For example, Hong Kong is an excellent region for building services. Additionally, the availability zone (AZ) refers to different physical locations, and selecting the appropriate one can affect performance and redundancy. If you choose to build all your servers in one AZ, your reliability will be compromised if that AZ experiences downtime. However, if you have your servers in different AZs, the servers may experience some latency when communicating with each other.
Another key consideration is the CPU architecture. The x86 architecture is ideal for those who want to run Windows applications, while the Kunpeng architecture, which is a proprietary chipset developed by Huawei, is generally less expensive. However, it may not be compatible with certain Windows applications, so it’s important to choose the appropriate CPU architecture based on your specific needs.
Finally, selecting the appropriate specifications for your server is crucial. The specifications will vary depending on the intended use of the server. For example, general computing is ideal for enterprise applications, memory-optimized specifications are suited for memory-intensive applications, while large memory specifications are ideal for applications like IMDb or SAP HANA. High-performance computing is ideal for activities such as gene sequencing. It is important to note that specifications may vary depending on the AZ, with some AZs offering fewer options than others.
In summary, when configuring your system, it is crucial to carefully consider the billing mode, region, AZ, CPU architecture, and specifications that are best suited to your specific needs. While this process may initially seem overwhelming, with time, you will become familiar with the options and select the appropriate configurations with greater ease.
When choosing a server, the specifications are crucial. You need to consider the size of the server you want, which includes factors such as CPU, memory, and assured bandwidth maximum barrier. These specifications vary depending on the type of application environment you are running, the amount of bandwidth your application needs, and the number of packets that come in and go out.
For general computing, you can choose a server with two virtual CPUs and 4GB memory, which is suitable for most purposes. However, if you require more advanced computing power, you may need to consider higher-performance servers with different CPUs.
The bandwidth is another critical factor to consider when selecting a server. The higher the server’s size, the higher the bandwidth, which is measured in assured bandwidth and maximum burstable bandwidth. Additionally, you need to pay attention to the CPU and the number of packets per second.
Once you have chosen your server specifications, you can then move on to selecting the image. The image represents the template or operating system of your server. You can choose from public, private, shared, or marketplace images. Public images are provided by the cloud environment developer, while private images are created by you. Shared images are shared from another account to you, while marketplace images are offered by third-party vendors.
In summary, when selecting a server, you should carefully consider your application’s specific requirements and choose a server that meets those requirements. The larger the server, the more computing power and bandwidth you will have, but also the higher the price. With the right server and image, you can ensure that your application runs smoothly and efficiently.
When using Huawei’s marketplace, you can find a variety of vendors or partners that offer images for different services such as database and caching with MongoDB, big data environments with TensorFlow, and security services including bastion hosts. To access these images, you simply need to browse the marketplace and select the ones you need. If you don’t find what you need, you can use a public image or send a request to us.
To start configuring your instance, you will first need to choose a system disk if you want to install an operating system, and data disks if you want to add more storage. After selecting your disks, you can proceed to configure your network. You can choose a Virtual Private Cloud (VPC) and subnets, and assign IP addresses manually if needed. Remember to plan your IP address range carefully to avoid conflicts with other networks.
To control access to your instance, you can use security groups, which act like firewalls. You can permit incoming and outgoing traffic based on specific port numbers and protocols. Be cautious and only allow the traffic you need to minimize security risks.
If you need a public IP address, you can choose to allocate an Elastic IP (EIP) for your instance. You can select the type of EIP, such as Dynamic BGP or Premium BGP, and the bandwidth or traffic you require. Note that higher bandwidth and premium options come with a higher price tag.
Lastly, you can configure your instance settings, including your login mode, key pair, and password. Keep in mind that password-based authentication may not be the most secure option.
Building an instance on Huawei’s platform can be straightforward and user-friendly, even for those new to the process. However, it’s essential to plan carefully and configure your instance securely to ensure reliable performance and protect your data.
Are you interested in backup and recovery, as well as a feature called anti-affinity? Anti-affinity ensures that each time you deploy a server, it is not placed on the same physical server as another server in the same group. By implementing anti-affinity, every server deployed in a group will be placed on different physical servers, which means that if one physical server goes down, your services will still be recovered without any downtime. In contrast, if you don’t have anti-affinity, there is a chance that multiple servers may be deployed on the same physical server, causing your services to be unavailable if that physical server goes down.
In addition to anti-affinity, there are advanced configuration options such as tags and agencies, which allow managed services to be provided. If you choose to create a managed service environment, you will need to provide an agent to your vendor. Once you have chosen the options you need, you can agree to the service level agreement and image disclaimer, and then submit your request to create a server.
Creating a server is not difficult, provided that you know what you want and need. With proper planning, you can have a server up and running in less than five minutes. Once your server is up, you will receive an IP address and a public ID. You can also remotely log in to the server using the VNC feature.
This is just a brief overview of the server creation process in the cloud environment. In the future, we can explore this topic further. However, for now, this concludes our discussion.