摘要:
A method for power-efficiency management in a virtualized cluster system is disclosed, wherein the virtualized cluster system comprises a front-end physical host and at least one back-end physical host, and each of the at least one back-end physical host comprises at least one virtual machine and a virtual machine manager for managing the at least one virtual machine. In the method, flow characteristics of the virtualized cluster system are detected at a regular time cycle, then a power-efficiency management policy is generated for each of the at least one back-end physical host based on the detected flow characteristics, and finally the power-efficiency management policies are performed. The method can detect the real-time flow characteristics of the virtualized cluster system and make the power-efficiency management policies thereupon to control the power consumption of the system and perform admission control on the whole flow, thereby realizing optimal power saving while meeting the quality of service requirements, so that a virtualized cluster system with high power-efficiency can be provided.
摘要:
A method and apparatus for power-efficiency management in a virtualized cluster system. The virtualized cluster system includes a front-end physical host and at least one back-end physical host, and each of the at least one back-end physical host comprises at least one virtual machine and a virtual machine manager. Flow characteristics of the virtualized cluster system are detected at a regular time cycle, then a power-efficiency management policy is generated for each of the at least one back-end physical host based on the detected flow characteristics, and finally the power-efficiency management policies are performed. The method can detect the real-time flow characteristics of the virtualized cluster system and make the power-efficiency management policies thereupon to control the power consumption of the system and perform admission control on the whole flow, thereby realizing optimal power saving while meeting the quality of service requirements.
摘要:
A method for rejecting received SIP messages which are redundantly retransmitted used in a Session Initiation Protocol (SIP) application architecture. In one embodiment, the method comprises: forwarding all the messages transmitted from a SIP application server to a SIP terminal, and dropping all the received SIP messages redundantly retransmitted from the SIP terminal to the SIP application server according to the transmission direction of said SIP messages; and correspondingly replying to the dropped SIP messages according to the message type thereof. The present invention also provides a corresponding retransmission rejection device and a network device comprising the same, and further provides a method and device for rapidly rejecting redundantly retransmitted SIP messages in the case where the memory space is limited. The present invention can be applied to a network device such as a SIP stateless proxy at a front end of a SIP application server cluster.
摘要:
A method, apparatus and system for processing of a series of service messages. In one embodiment, the series of service messages is formed by a plurality of service messages, and at least two server instances provide real-time processing of the service messages and store the information provided by the service messages and results obtained from processing the service messages. The method includes the steps of: determining an activity that affects real-time processing of the service messages will occur in at least one server instance of the server instances; and dispatching subsequent service messages among the service messages to other server instance(s) other than the at least one server instance, so as to ensure that said activity will not affect message processing delay.
摘要:
A method for rejecting received SIP messages which are redundantly retransmitted used in a Session Initiation Protocol (SIP) application architecture. In one embodiment, the method comprises: forwarding all the messages transmitted from a SIP application server to a SIP terminal, and dropping all the received SIP messages redundantly retransmitted from the SIP terminal to the SIP application server according to the transmission direction of said SIP messages; and correspondingly replying to the dropped SIP messages according to the message type thereof. The present invention also provides a corresponding retransmission rejection device and a network device comprising the same, and further provides a method and device for rapidly rejecting redundantly retransmitted SIP messages in the case where the memory space is limited. The present invention can be applied to a network device such as a SIP stateless proxy at a front end of a SIP application server cluster.
摘要:
A method, apparatus and system for processing of a series of service messages. In one embodiment, the series of service messages is formed by a plurality of service messages, and at least two server instances provide real-time processing of the service messages and store the information provided by the service messages and results obtained from processing the service messages. The method includes the steps of: determining an activity that affects real-time processing of the service messages will occur in at least one server instance of the server instances; and dispatching subsequent service messages among the service messages to other server instance(s) other than the at least one server instance, so as to ensure that said activity will not affect message processing delay. The impact of the activity on real-time processing of service messages can be eliminated, while the dramatic decrease of the throughput of service messages, which will be occurred if other methods for improving service message processing delay are employed, can be avoided.
摘要:
A method and device for QoS control for presence servers and a presence system supporting QoS control. The method for providing QoS capability for a presence server includes the steps of: receiving SIP (Session Initiation Protocol) messages; classifying said SIP messages based on a QoS policy; and delivering the classified SIP messages after a predetermined time corresponding to the QoS policy.
摘要:
The invention provides an information polling method, apparatus and system thereof, wherein a message gateway receives an information polling request from an application in an application client, and according to the received information polling request, the message gateway acquires at least one message requested by the application and calculates a reference polling cycle of the application, and sends the acquired message requested by the application and the reference polling cycle to the application. The application determines a polling cycle of a next poll according to a load condition of the application client and the received reference polling cycle. According to the invention, efficiency and quality of the information polling is improved, while the system is ensured to run normally, reducing the cases of message blocking and system breakdown which may occur in the existing situations.
摘要:
To reduce traffic between multiple servers in a presence system and improve the performance of the presence system, a method and device is provided for user-management in a presence system, and a presence system. The presence system comprises multiple presence servers communicating through a network, and said device for user-management comprises a combining means for combining multiple user sets among which there are the greatest correlations into a user set to be allocated; and an allocating means for allocating said user set to be allocated to one presence server which can contain said user set to be allocated. Through allocating multiple user sets among which there are the greatest correlations to a same presence server, communications among users in the user set are performed within a same presence server, reducing occupations for network resources and improving performance of the presence system.
摘要:
A method and an apparatus of processing Session Initiation Protocol (SIP) messages based on multiple cores. The method comprises: pre-parsing a received SIP message to obtain an identifier of the SIP message; dispatching SIP messages with the same identifier to the same core; and processing, on each core, SIP messages dispatched to the core by using related dedicated resources. The present invention relieves the resource contention and thus improves the utilization efficiency of computing resources of cores.