移动通讯设备高可用系统中的数据存储技术
摘要
在移动通讯中，高可用性系统有着广泛的应用，而数据对于移动通讯业务来说也是相当重要的。因此在实现移动通讯的系统时，需要尽量的提高系统的可用性和数据存储的安全性。提高系统的可用性最常见的方法是进行系统备份，系统备份一般采用的方法有2N备份和N+1备份，系统备份从层面上来说又分为硬件层面和软件层面的系统备份；数据存储的安全性也有很多方法，由于电信系统中数据交换速度快，数据一致性的要求高，所以数据存储需要软件层面和硬件层面的双重保障。
本文着重阐述项目系统中对高可用性的研究和部分的高可用性的实现，项目系统采用了Cpx8216框实现基站的通信业务功能，为了保证整个系统的高可用性，采用了主、备用的2N备份模式，采用热重启的工作模式，在系统局部采用的N+1备份方式，从硬件上保证了系统的可用性；在软件设计上配合硬件的备份模式，保证整个系统的可用性，同时还采用合理的软件检测控制，调度设备，进一步提高系统的可用性。
本文还研究了在高可用性系统中的数据存储，在实际项目中采用RAID＋DBMS的方式实现了数据存储。RAID的采用主要是从硬件上保证了存储数据的安全，减少由于硬件故障导致的数据丢失；而DBMS则是从软件上保证了数据的安全和可用性，避免对数据进行非法读、写等操作，同时采用数据库系统还能在一定程度上保证数据的一致性，也就是数据的有效性。为了适应电信业务的数据交换速度必须选择合适的数据库管理系统，同时还要合理的设计数据库才能最大限度的提高系统的可用性和数据存储的安全性。
实际项目同时考虑硬件设备和软件设计两方面的因素，不仅保证了系统有相当地可用性，数据存储也完全满足了系统的功能需求，使整个系统的性能有了相当大的提高。
High Availability System (HA) and Safe Data Storage are of great importance and widely applied in mobile communications. To improve the availability of HA, System Redundancy, which can be classified into 2N Redundancy and N+1 Redundancy, or Hardware Redundancy and Software Redundancy, is the most common approach. To improve the security of Data Storage, there are many approaches, and because of the high data exchange speed and the high consistency requirement in Data Storage in Telecom System, the dual guarantees respectively based on Software Layer and Hardware Layer are required.
In the paper, the realization methods of High Availability (HA) and Data Storage in HA System of my project, which are based on Cpx8216 base station system frame, are proposed. To guarantee the HA of the system, in Hardware Layer, the 2N Redundancy mode for the overall system, N+1 Redundancy mode for the partial system and Hot Restart work mode are applied. And in Software Layer, software designs corresponding to the hardware Redundancy mode and for the availability of the overall system, and suitable software for checking, controlling and arranging the equipments, are adopted.
In the paper, Data Storage in HA System, which is realized by RAID＋ DBMS mode in the practical project, is studied , as well . Here, RAID is to guarantee the security of the storage data and reduce the data losses caused by hardware failures in Hardware Layer, DBMS is to guarantee the security and availability of the data, avoid the illegal data operations, such as illegal reading and illegal writing, and the application of the database system is to guarantee the consistency, namely availability, of the data. Moreover, the suitable DataBase
Manager System is required for the high data exchange speed in the telecom operations and the suitable database design is required to improve the availability of the system and security of the Data Storage.
In the practical project, the hardware equipments and the software design are both considered, that not only guarantee the availability of the system, but also accord with the requirements of the system functions for Data Storage, and, as a result, the overall performances of the system are improved to a extent.
本文着重阐述项目系统中对高可用性的研究和部分的高可用性的实现，项目系统采用了Cpx8216框实现基站的通信业务功能，为了保证整个系统的高可用性，采用了主、备用的2N备份模式，采用热重启的工作模式，在系统局部采用的N+1备份方式，从硬件上保证了系统的可用性；在软件设计上配合硬件的备份模式，保证整个系统的可用性，同时还采用合理的软件检测控制，调度设备，进一步提高系统的可用性。
本文还研究了在高可用性系统中的数据存储，在实际项目中采用RAID＋DBMS的方式实现了数据存储。RAID的采用主要是从硬件上保证了存储数据的安全，减少由于硬件故障导致的数据丢失；而DBMS则是从软件上保证了数据的安全和可用性，避免对数据进行非法读、写等操作，同时采用数据库系统还能在一定程度上保证数据的一致性，也就是数据的有效性。为了适应电信业务的数据交换速度必须选择合适的数据库管理系统，同时还要合理的设计数据库才能最大限度的提高系统的可用性和数据存储的安全性。
实际项目同时考虑硬件设备和软件设计两方面的因素，不仅保证了系统有相当地可用性，数据存储也完全满足了系统的功能需求，使整个系统的性能有了相当大的提高。
High Availability System (HA) and Safe Data Storage are of great importance and widely applied in mobile communications. To improve the availability of HA, System Redundancy, which can be classified into 2N Redundancy and N+1 Redundancy, or Hardware Redundancy and Software Redundancy, is the most common approach. To improve the security of Data Storage, there are many approaches, and because of the high data exchange speed and the high consistency requirement in Data Storage in Telecom System, the dual guarantees respectively based on Software Layer and Hardware Layer are required.
In the paper, the realization methods of High Availability (HA) and Data Storage in HA System of my project, which are based on Cpx8216 base station system frame, are proposed. To guarantee the HA of the system, in Hardware Layer, the 2N Redundancy mode for the overall system, N+1 Redundancy mode for the partial system and Hot Restart work mode are applied. And in Software Layer, software designs corresponding to the hardware Redundancy mode and for the availability of the overall system, and suitable software for checking, controlling and arranging the equipments, are adopted.
In the paper, Data Storage in HA System, which is realized by RAID＋ DBMS mode in the practical project, is studied , as well . Here, RAID is to guarantee the security of the storage data and reduce the data losses caused by hardware failures in Hardware Layer, DBMS is to guarantee the security and availability of the data, avoid the illegal data operations, such as illegal reading and illegal writing, and the application of the database system is to guarantee the consistency, namely availability, of the data. Moreover, the suitable DataBase
Manager System is required for the high data exchange speed in the telecom operations and the suitable database design is required to improve the availability of the system and security of the Data Storage.
In the practical project, the hardware equipments and the software design are both considered, that not only guarantee the availability of the system, but also accord with the requirements of the system functions for Data Storage, and, as a result, the overall performances of the system are improved to a extent.
引文
[1] W.David Shambroom, charles River Data Systems， Inc. "Use of Protocol Validation and verification Techniques in the design of a Fault-Tolerant Computer Architecture", 0731-3071/93 $3.00 @ 1993.IEEE
[2] Seungjae Han and Kang G. Shin, Fellow, IEEE. "Experimental Evaluation of Behavior-Based Failure-Detection Schemes in Real-Time Communication Networks" IEEE Trans. Parallel and Distributed System, VOL. 10, NO. 6, JUNE 1999
[3] Ravi K. Iyer Center for Reliable & High-Performance Computing Coordinated Science labo`ratory University of Illinois at Urbana-Champaign and A. Avizienis Computer Science Department 4731 Boelter Hall University of California, Los Angeles "COTS Hardware and Software in High-Availability Systems "
[4] P.K. Chande, Senior Member IEEE and S.V. Tokekar, Member IEEE "Expert-Base Maintenance: A study of Its Effectiveness" IEEE Trans. On Reliability, Vol. 47 No. 1, March 1998
[5] Torbjrnsen, ClustRa AS Svein-Olaf Hvasshovd NTNU & Maitrayi Sabaratnam, Norwegian University of Science & Technology NTNU，"Evaluating the Effectiveness of Fault Tolerance in Replicated Database"
[6] S. Bratsberg, Gruvlen, S. Hvasshovd, B. Munch, and Torbjrnsen. "Providing a Highly Available Database by Replication and Online Self-Repair" International Journal of Engineering Intelligent Systems, 4(3):131-139, 1996.
[7] Ravi Mukkamala "Storage Efficient and Secure Replicated Distributed Databases"，IEEE Trans. Knowledge and Data Eng., Vol. 6, no. 2,April 1994.
[2] Seungjae Han and Kang G. Shin, Fellow, IEEE. "Experimental Evaluation of Behavior-Based Failure-Detection Schemes in Real-Time Communication Networks" IEEE Trans. Parallel and Distributed System, VOL. 10, NO. 6, JUNE 1999
[3] Ravi K. Iyer Center for Reliable & High-Performance Computing Coordinated Science labo`ratory University of Illinois at Urbana-Champaign and A. Avizienis Computer Science Department 4731 Boelter Hall University of California, Los Angeles "COTS Hardware and Software in High-Availability Systems "
[4] P.K. Chande, Senior Member IEEE and S.V. Tokekar, Member IEEE "Expert-Base Maintenance: A study of Its Effectiveness" IEEE Trans. On Reliability, Vol. 47 No. 1, March 1998
[5] Torbjrnsen, ClustRa AS Svein-Olaf Hvasshovd NTNU & Maitrayi Sabaratnam, Norwegian University of Science & Technology NTNU，"Evaluating the Effectiveness of Fault Tolerance in Replicated Database"
[6] S. Bratsberg, Gruvlen, S. Hvasshovd, B. Munch, and Torbjrnsen. "Providing a Highly Available Database by Replication and Online Self-Repair" International Journal of Engineering Intelligent Systems, 4(3):131-139, 1996.
[7] Ravi Mukkamala "Storage Efficient and Secure Replicated Distributed Databases"，IEEE Trans. Knowledge and Data Eng., Vol. 6, no. 2,April 1994.