Data sending method and device

Abstract

The invention provides a data transmitting method and a data transmitting device. The method includes the following steps that when a data transmitting process is dispatched, data to be transmitted are obtained from a core framework, the data are transmitted to a buffering queue corresponding to channels, all channels included in one channel group, corresponding to one physical port, of multiple channel groups are polled, if data exist in the buffering queue corresponding to the channels, the data in the buffering queue are sequentially transmitted to hardware FIFO, and the hardware is used for transmitting the data; after all the channels in the channel group are polled, the position information of the channel group is recorded, and the data transmitting process is finished, so that when the data transmitting process is dispatched next time, all channels in the next channel group are polled according to the recorded position information of the channel group. According to the data transmitting method and the data transmitting device, by dispatching and polling one channel group of the same physical port, the polling and dispatching efficiency and the data transmitting efficiency of multiple channels of the physical port are improved, and the performance of a router is improved.

Description

technical field

[0001] The invention relates to communication technology, in particular to a data sending method and device. Background technique

[0002] Existing routers generally use a fast-forward mode to send and receive data, and the fast-forward mode performs round-robin scheduling on channels in units of physical ports. For channelized interfaces, one physical interface has multiple channels. For example, the E1 interface of the WAN interface with cable as the transmission medium has at most 31 channels, and the CPOS interface of the WAN interface with optical fiber as the transmission medium has at most 64*31 channels. When using the fast-forward mode to perform round-robin scheduling on a physical port, multiple channels of a physical port need to be processed, that is, round-robin scheduling drives buffer queues corresponding to multiple channels for data transmission and reception.

[0003] The existing technology uses the channel bitmap to perform two queries, ...

Examples

Embodiment 1

[0021] figure 1 It is a flow chart of the data sending method provided by Embodiment 1 of the present invention. The embodiment of the present invention is suitable for performing data sending and receiving processing on multiple channels belonging to one physical port in the fast-forward mode, that is, polling and scheduling the channels in units of physical ports. When receiving and processing data on multiple channels of the physical port, the hardware has already received the data into the first-in-first-out queue (First Input First Output, FIFO for short), and each round-robin scheduling only needs to extract data from the FIFO and upload it to the upper layer for forwarding Or local processing, so the data receiving processing for one physical port will not affect the scheduling processing for other physical ports, but when all channels of a physical port have data to send at the same time, all channels of a physical port need to be rotated Inquiry and scheduling proces...

Embodiment 2

[0028] On the basis of Embodiment 1, the data to be sent is acquired from the core framework, and before the data is sent to the buffer queues of the corresponding channels, the buffer queues corresponding to all channels of the same physical port are reset. The data to be sent is obtained from the core framework, and if the obtained data is not empty, the data is sent to the buffer queue of the corresponding channel.

[0029] Before obtaining the data to be sent from the core framework, first reset the buffer queues corresponding to all channels of the same physical port, specifically, reset the hardware state of the channel, or clear the data in the buffer queue corresponding to the channel.

[0030] If the data to be sent obtained from the core framework is not empty, the data is sent to the buffer queue of the corresponding channel, and if the number of the data exceeds the length of the buffer queue, the data exceeding the length of the buffer queue is discarded.

[0031]...

Embodiment 3

[0034] figure 2It is a schematic diagram applicable to the data sending method provided by Embodiment 3 of the present invention. On the basis of the above embodiment, the sequentially sending the data in the buffer queue to the hardware FIFO, and sending the data by hardware includes sending one piece of the data in the buffer queue to the hardware FIFO, Make the hardware send the data from the hardware FIFO in a first-in-first-out manner; or send the data in a plurality of buffer queues to the hardware FIFO, so that the hardware sequentially sends the data from the hardware FIFO in a first-in-first-out manner Send the data.

[0035] Such as figure 2 As shown in , after sending the data obtained from the core framework to the buffer queue of the corresponding channel, the channel is polled, and if there is data to be sent in the buffer queue corresponding to the channel, the data in the buffer queue is sequentially sent to Hardware FIFO; for example, a channel group cont...

technical field

[0001] The invention relates to communication technology, in particular to a data sending method and device. Background technique

[0002] Existing routers generally use a fast-forward mode to send and receive data, and the fast-forward mode performs round-robin scheduling on channels in units of physical ports. For channelized interfaces, one physical interface has multiple channels. For example, the E1 interface of the WAN interface with cable as the transmission medium has at most 31 channels, and the CPOS interface of the WAN interface with optical fiber as the transmission medium has at most 64*31 channels. When using the fast-forward mode to perform round-robin scheduling on a physical port, multiple channels of a physical port need to be processed, that is, round-robin scheduling drives buffer queues corresponding to multiple channels for data transmission and reception.

[0003] The existing technology uses the channel bitmap to perform two queries, ...

Examples

Embodiment 1

[0021] figure 1 It is a flow chart of the data sending method provided by Embodiment 1 of the present invention. The embodiment of the present invention is suitable for performing data sending and receiving processing on multiple channels belonging to one physical port in the fast-forward mode, that is, polling and scheduling the channels in units of physical ports. When receiving and processing data on multiple channels of the physical port, the hardware has already received the data into the first-in-first-out queue (First Input First Output, FIFO for short), and each round-robin scheduling only needs to extract data from the FIFO and upload it to the upper layer for forwarding Or local processing, so the data receiving processing for one physical port will not affect the scheduling processing for other physical ports, but when all channels of a physical port have data to send at the same time, all channels of a physical port need to be rotated Inquiry and scheduling proces...

Embodiment 2

[0028] On the basis of Embodiment 1, the data to be sent is acquired from the core framework, and before the data is sent to the buffer queues of the corresponding channels, the buffer queues corresponding to all channels of the same physical port are reset. The data to be sent is obtained from the core framework, and if the obtained data is not empty, the data is sent to the buffer queue of the corresponding channel.

[0029] Before obtaining the data to be sent from the core framework, first reset the buffer queues corresponding to all channels of the same physical port, specifically, reset the hardware state of the channel, or clear the data in the buffer queue corresponding to the channel.

[0030] If the data to be sent obtained from the core framework is not empty, the data is sent to the buffer queue of the corresponding channel, and if the number of the data exceeds the length of the buffer queue, the data exceeding the length of the buffer queue is discarded.

[0031]...

Embodiment 3

[0034] figure 2It is a schematic diagram applicable to the data sending method provided by Embodiment 3 of the present invention. On the basis of the above embodiment, the sequentially sending the data in the buffer queue to the hardware FIFO, and sending the data by hardware includes sending one piece of the data in the buffer queue to the hardware FIFO, Make the hardware send the data from the hardware FIFO in a first-in-first-out manner; or send the data in a plurality of buffer queues to the hardware FIFO, so that the hardware sequentially sends the data from the hardware FIFO in a first-in-first-out manner Send the data.

[0035] Such as figure 2 As shown in , after sending the data obtained from the core framework to the buffer queue of the corresponding channel, the channel is polled, and if there is data to be sent in the buffer queue corresponding to the channel, the data in the buffer queue is sequentially sent to Hardware FIFO; for example, a channel group cont...

Images