When data center owners build cross-data center network interconnections, they mainly consider issues such as large bandwidth, low latency, high density, fast deployment, easy operation and maintenance, and high reliability. At present, the mainstream large-bandwidth OTN technology is mainly controlled by several large-scale telecom equipment manufacturers (chips are discussed separately), such as Huawei, ZTE, and Aran. The main customers they face are traditional telecom operators, so the product features of OTN are mainly designed to meet the service characteristics of these operators. Because of this, OTN is increasingly used in DCI network applications in the Internet industry. Many discordant problems.
The characteristics of OTN equipment are the same problems encountered by DCI, rich business overhead, strong OAM capability of the network, scheduling and multiplexing capabilities of different granular bandwidths, line fault tolerance under long-distance conditions, and the use of low-voltage direct current. Features such as low utilization rate of power consumption of equipment.
1. Rich business spending capabilities require O&M personnel to be more professional, rely more on technical support from manufacturers, and have a more closed technology.
2. Powerful OAM capabilities, inconsistent standards, more difficult and more independent interconnection with cross-networks, and useless functions also bring more transmission and operation costs to the DCI network.
3. Different granule scheduling capabilities make the service encapsulation frame structure more complicated and more nested bytes.
4. The long-distance line fault tolerance makes the FEC algorithm complex, consumes more overhead and takes longer to process
5. The 48V-DC power supply mode of OTN equipment is different from the standard 19-inch 220V-AD (or 240V-DC) cabinets used in most data centers. The installation is complicated and requires power transformation in the computer room
6. The traditional OTN equipment has a large frame, which is not suitable for installation in standard cabinets, and the capacity density is not high. Later expansion is troublesome and requires cabinets to be moved or rebuilt.
At present, our DCI network is mainly to provide pipelines for cross-data center data. The main characteristics of the business model are: unified and single bandwidth granularity requirements, large bandwidth, cross-data center services (especially multi-active IDC, big data services) ) has low latency requirements and high requirements for network stability; at the same time, due to the lack of relevant professional and technical personnel in the Internet industry, the operation and maintenance of DCI networks needs to be “simple” “simple” and “simple” – important things are said three times (where Isn’t it a kind of network?); the explosive development of the Internet makes the construction and expansion cycle requirements shorter (operators’ OTN expansion cycle is generally half a year to one year, while the Internet’s own DCI expansion requirements are 1 to 3 months), Therefore, it is necessary to compress the time in all aspects.
Therefore, OTN provides a usable solution for DCI, but OTN is by no means the most suitable solution for DCI. Now that the DCI network is booming, there is an increasing need for some suitable solutions to solve problems ranging from cost to construction, operation and maintenance. various problems encountered. And these problems are nothing more than the six requirements of the DCI network (large bandwidth, low latency, high density, fast deployment, easy operation and maintenance, and high reliability):
1. Large bandwidth, DCI transmission network does not have rich types of granules like operators, the bandwidth granules of DCI transmission network are simpler, currently commonly used 10G or 100G, in the future 200G/400G, etc., so there is no need to do it with large bandwidth Bandwidth at other granularities. Due to the fact that the distance range of the DCI transmission network is generally not too long, using the 400G system based on 200G PM-16QAM dual-carrier modulation, the transmission distance without electrical relay can be about 500 kilometers (PM-64QAM is about 200 kilometers) , so that the metro backbone transmission of DCI will not be limited by distance.
2. Low latency, DCI business requirements, especially when cloud computing is used to pool resources and multi-active data centers, the latency is calculated at the microsecond level, so the data transmission time should be as short as possible, wishing to exceed the speed of light. Eliminate unnecessary data processing work and reduce signal transmission paths. For example, by removing the SD-FEC function used by 100G OTN, a single back-to-back can save 200 microseconds, and removing the cross-level OTN encapsulation can save tens of microseconds, and rationally use the hubspoke topology for key services to ensure the shortest path. Of course, it can also cooperate with MPLS and QOS at the IP level to try to ensure that the delay at the data forwarding level is also better.
3.High density, a single U, or 2U, can achieve a bandwidth of up to T, DWDM optical layer and electrical signal layer decoupling, improve the density interface of the device, and reduce the size of the optical module. For example, the use of QSFP28 optical modules can ensure that the 100G access capability of a single device is greatly improved, and the use of CFP2 daylighting modules on the line side can ensure that the transmission bandwidth of the overall device is improved, 1U can be 1.6T, 3.2T. At present, many related products have appeared in the world, such as ADVA, coriant, ciena and other companies. Of course, domestic Huawei has also launched related 902 products. However, as of the completion of this article, it seems that the network access test of the Ministry of Industry and Information Technology has not been completed. High density will cause high power consumption and heat dissipation. Therefore, the original OTN heat dissipation method of left and right air in and out, and up and down air in and out should be discarded.
Device cooling requirements.
4. Fast deployment, using the current standardized IDC 19-inch rack, similar to the form of mainstream servers, using AC-220V for direct power supply, eliminating the need for power and cabinet transformation, and realizing that the goods can be put on the shelves immediately after they arrive in the computer room, and can be configured after plugging in the power supply. business, and do a good job in standardized acceptance work to achieve rapid deployment.
5. Easy operation and maintenance, DCI business model requirements, the distance across data centers will not be very far, and complex management overhead, OAM and other functions are not necessary to play a big role in this scenario, and complex processing is also reduced It improves data transmission efficiency, improves data processing time, and has higher requirements for technology and is more closed. Connecting signals directly via Ethernet eliminates the complicated overhead of OTN, so traditional IP network engineers can operate and maintain the DCI system. After combining new northbound interfaces such as YANG model, REST API, and netconf, DCI transmission equipment management and IP network equipment management are developed with the same interface, so as to better carry out unified platform-based centralized network management.
6. High reliability, multi-physical routing and protection technologies that are not aware of the upper layer will continue to play a role in the DCI transmission network. Failures at the underlying link level should not have any impact on services unless they are completely interrupted. Perception or impact, whether it is protection switching, link jitter, delay increase, etc.
Post time: Feb-13-2023