Cmljhxhyy system architecture plan

I. System Architecture Design Topology

Topology description:

1. The above hardware is not limited to brands and models. As a system architecture design, it aims to highlight the configuration features and effects.

2. The above software components depend on existing resources and application system implementation resources. To ensure long-term effectiveness and O & M efficiency, software functions are provided based on implementation services. Users do not need to purchase related software. It is recommended to use the O & M Support Service to facilitate the high maintainability of system construction.

3. For a detailed list of the software and implementation technologies used, see the attachment.

Ii. design purpose and problems solved

1. The San network design model fully considers the 5-year planned usage of the new building. After the "New Building Network Construction", "server and application architecture construction such as PACS" is solved.

2. multi-dimensional multi-level High Availability design is a highlight, the integrated design covers failover in multiple application scenarios (storage, San network, server hardware, operating system cluster, virtualization cluster, application server cluster, Database Cluster, database image, and data synchronization )., real-time data change, Application Deployment switch ). It solves the problem that the system does not have a single point of failure and can be fault-tolerant for any fault. It can restore applications within 30 minutes, recover databases within 3 minutes, and save data images within 1 minute, second-level almost real-time his form-level data synchronization.

3. upgrade the O & M and Emergency Response of a single hardware and application to a more reliable and effective "Overall O & M assurance service", one-step solution to the problems originally caused by multi-party joint debugging, including: spare parts, change implementation, fault diagnosis, emergency response, performance tuning, and application optimization. O & M assurance objectives: high availability, high reliability, high performance, low risk, low O & M costs, and low management costs.

4. Make full use of old devices, introduce a third-party archive backup mechanism, and make full use of old devices.

Iii. Implementation results

1. the informatization of the new building will fully comply with the informatization construction standards of the top three hospitals.

2. His/PACS/EMR/LIS/herp have laid a solid foundation for system architecture design.

3. Make full use of the old data. After fine-tuning through the virtualization platform, it can be fully used for archiving and third-party backup, creating a three-center backup system in two locations to achieve intensive equipment management.

4. Do not worry about the risk of service interruption caused by unexpected downtime. The system recovery time can be controlled within seconds (Database fault) to a maximum of 30 minutes based on the severity of the fault (full failure of system storage hardware and software ). This recovery indicator can be used as a standard for disaster recovery drills and a benchmark for the medical industry.

5. the informatization construction level of O & M assurance has been upgraded to a new level to achieve refined management of "Three highs, three lows: high Availability, high reliability, high performance, low risk, low O & M costs, and low management costs.

Iv. unrecommended design methods and reasons

1. After thorough research and application compatibility analysis, and data volume growth analysis. We do not recommend that you deploy a platform based on virtualization technology on a large scale. It is only suitable for a small number of integrated reuse and compatibility tests. We recommend that you use a mature open-source free KVM/ESX method, there is no need to purchase new virtualization software such as VMWare, which is cost-effective and creates maintenance risks.

2. We do not recommend that you use a single-controller storage product, which may result in a controller failure and cause system unavailability. We recommend that you use a storage product with multiplexing technology.

3. We do not recommend that you use a single raid or a hybrid disk combination to create a storage. Considering the large demand for His/PACS storage, see the previous technical solutions for detailed differences. We recommend that you purchase storage specifications that support "Data layering" technology.

4. private technologies or incompatible storage management protocols are not recommended. Considering the advantages of the old architecture and the future expansion and upgrade of PACS and other systems, the selected basic storage device must support heterogeneous storage management, that is, multiple brands and models can all use a unified interface management and unified technical maintenance.

5. because many software-layer cluster technologies have many matching details with specific software versions and technologies, such as hybrid clusters built with different database versions, how does the hardware failover mechanism match the OS failover mechanism, his application code-level failover trigger Condition Monitoring granularity and so on. Therefore, it is recommended that the system architecture implementation supplier be responsible for all hardware, software and services and instruct the hardware supplier to implement them. Do not disconnect the supply and implementation of equipment, which leads to division of rights and responsibilities and mutual pressure, which is not conducive to the success of the project.

V. Related Technical list

To ensure the implementation of the above application-level software, the involved software and related technical content are as follows:

1. Operating System Cluster (for failover when the operating system fails) the operating system has its own license, which is free of charge.

2. Operating System Network Server Load balancer cluster (used for non-fault web systems, but the performance of OA and other Web systems is slow, resulting in user dissatisfaction) the operating system comes with a license, free of charge.

3. SQL Server database failover cluster (used for database instance failover in the database environment) operating system comes with a license, free of charge.

4. SQL Server database image cluster (Real-Time instance data switching in the user database environment) comes with a license and is free of charge.

5. SQL Server database continuous data replication (Real-time table-level fault recovery in the user database environment) comes with a license and is free of charge.

6. vsphere VMware ha implementation (Virtual Machine-level failover) is implemented using open-source KVM, xen, and vsphere hypervisor. Product, KVM, xen, and vsphere hypervisor are mature open-source free products.

7. vsphere VMware ft implementation (Real-Time failover of key data module virtual machines) license is the same as above

8. vsphere VMware Dr implementation (Virtual Machine-level backup-based disaster recovery implementation) license is the same as above

9. vsphere esxi fault recovery design (esxi is the core of virtualization, and esxi failure will make all virtual machines unavailable, so the esxi fault recovery solution is particularly important.

10. TSM for dB and TSM for San implementation (Backup design and Disaster Tolerance software) IBM commercial software. We recommend that you purchase the O & M assurance service provided by the implementation supplier, which is provided by the supplier, you do not have to purchase it on your own.

11. CDC for SQL Server implementation (real-time data synchronization software) IBM commercial software. We recommend that you purchase the O & M assurance service provided by the implementation supplier, which is provided by the supplier and does not have to be purchased by yourself.