How to deal with cloud burst architecture? Four different ways to melancholy you.

"TechTarget China original" although most CIOs like hybrid cloud solutions, the reality is quietly encountering some annoying little problems-such as the lack of investment in some of the United States and EU telecommunications services fibre-optic connections. Welcome to the Infernal Web experience of cloud bursting architecture. The lack of bandwidth between public and private clouds makes cloud outbreaks and agility require a theoretical concept of repositioning it workloads. In addition, the cost of migrating data and system configuration in the cloud burst architecture is much larger than the bandwidth and storage access on the local area network (LAN). This network resource-related issue has led to two companies from NetApp and Verizon, which have established joint ventures to collaborate to transfer users to Verizon data centers to accelerate access to the public cloud. The problem is also the need for businesses that want cloud outbreaks to be faced with: transferring external data to an on-premises hybrid cloud. In essence, the bandwidth problem persists, but now the problem is in the local private part of the hybrid cloud, not the cloud provider. There are four ways to solve problems that may be encountered in cloud outbreaks. Abandon hybrid cloud--to public cloud corporate IT operations can be migrated to a common cloud. Some will sacrifice the Symbian-Okliss Act and the Health Insurance Circulation and Accountability Act, which are not simple questions. Hosting is acceptable in some way, because it means you have an IT infrastructure that is hosted in a compliant building, but the public cloud is a whole beast with a large number of tenants. In 2016, we can see that Amazon Web Services (AWS) and other cloud service providers provide dedicated systems and two solutions for long-term contracts, effectively creating the ability to meet regulatory requirements. Managed cloud services provide the necessary cloud burst architecture, coupled with fast storage access, to make the cloud truly flexible. The mechanism is long-term and ensures the stability of the desired configuration with absolute security for enterprise data and system access, while guaranteeing flexibility and capability during large bursts. Get a fast WAN another option is to connect the private and managed parts of the hybrid cloud via the fast WAN link of the telco operator. This allows the normal payload to use fast LANs, high-speed fiber-optic links to large public clouds, cloud bursting or workload rebalancing, which is much better than implementing an internal cloud.   From internal outbreak   Another potential cloud bursting architecture is to deploy everything in its own data center. Internal cloud outbreaks may require careful scheduling and frequent rebalancing, but this is not a very sophisticated automated process, so it is not easy to implement. Building a wall another way is to partition the data, which is part of the subset that stays in the public cloud forever. Consider such a schema if the data is transient, such as some big data classes, and the boundaries between public and private clouds can be dynamic. For example, a sensor data flow set from a retail store can be preceded by a user name before reaching the Cloud analytics platformRow, and then dynamically split according to the initial name. Choosing the right architecture for enterprise IT organizations in all the discussions, the first is to face economic problems. Is the public cloud cheaper than the on-premises cloud? Can hosting save money? These problems are far more complicated than the simple price problem. Migrating workloads between your own data center, hosting room, or public cloud, any solution can affect dynamic performance, so it wants to operate in a more stable environment rather than switching from one place to another. Security is also a problem, although public cloud providers such as AWS have invested heavily in anti-threat and tenant isolation measures. In general, the seminar shifted from being outsourced to the work load to some rumors and stereotypes about security. The cost is the next question. The complexity of running time factors is at the cost of consideration. Storage rhetorical question efficiency, virtual instance performance, and other operational-related factors-these calculations can affect IT manager judgment. If you implement the purchase of a dedicated server/instance on a public cloud, you should be able to directly mimic the configuration of the instance and even the internal server. The biggest problem may come from accessing the storage speed. Because the internal may be faster, because the data flow can be tuned to some extent. Private systems within the public cloud may be close to internal server performance. Access to a traditional hosting facility to a third-party vendor is as fast as on-premises operations, but a public cloud outbreak can cause some delay. A long-term commitment is required for the migration of any managed solution. A typical partnership contract locks the IT organization vendor for 1-2 years-a key decision. Although the result may be a cloud service, the buyer loses the right to shop around and gain advantage. The uplink quality of managed facility vendors is another factor that impacts cloud outbreaks. Quality is far more important than diversification choices, and the hoster may have multiple uplink options to connect to different data centers. In addition, disruption due to changes, migrations, and replication data, change workflows and processes, and audit governance are all elements required for any cloud-bursting architecture financial model. Telecom operators must start the long-delayed installation of urban fibre-optic connections, even for the long-suffering IT companies that need cloud bursts. The hybrid cloud will most likely exist in the cloud as a managed server, or fully managed, as long as the network bandwidth problem can be resolved. It's almost impossible to get back to a self-built data center. Not all discussions about cloud computing consider a hybrid structure. It is possible to migrate all it workloads to a public cloud or an all-private cloud architecture, which many IT organizations now operate. The agility of cloud outbreaks and server capacity planning requirements may increase costs, but they are still a few acceptable patterns for large-scale cloud users who do not accept hybrid cloud.

TechTarget China original content, original link: http://www.searchdatacenter.com.cn/showcontent_92335.htm
©techtarget China: http://www.techtarget.com.cn

How to deal with cloud burst architecture? Four different ways to melancholy you.