Does DC support the 802.1Q VLAN encapsulation technology?
Support.
How do DC users use the DC of DC providers?
First, the DC provider needs to establish NNI interconnection with the DC point of the cloud platform. The DC provider needs to go to the DC console and apply for a connection, and the DC delivery engineer will assist in completing the line connection and making NNI interconnection. During this process, there will be a connection ID under the DC provider's cloud platform account, in the format of dc-*******. After it is connected, users use your DC in a shared connection mode:
- The DC provider needs to prepare its own cloud account ID and connection ID (
dc-*******), as well as non-conflicting VLAN IDs and even interconnection IP information, and provide them to its users. - The DC user of the DC provider directly opens the cloud platform DC tunnel console , click Create+, select Shared Connection in Connection, enter the DC provider's account ID, connection ID, and VLAN ID in Advanced Configuration, and click Submit.
- The DC providers need to open a DC tunnel console in the cloud platform account, click Accept in the far right column of the DC tunnel applied for by the user.
- The cloud platform will automatically create a tunnel within a few minutes.
How can I choose the connection port type?
Confirm with your DC provider. The optical port module supports long-distance/short-distance connections. Confirm the port type again before you submit a connection application to the cloud platform.
Does it support multi-line CLB or active-standby relationship?
Support. DC Tunnel:
- BGP Routing:
- CLB (Active-active): Regardless of whether multiple lines are connected to the same DC point, as long as the local data center under the cloud publishes equal-cost routes to the same DC gateway on the cloud, the traffic sent from the DC gateway on the cloud to the local data center under the cloud will be CLB;
- Primary/Replica (Active-standby): Regardless of whether multiple lines are connected to the same DC point, users need to press multiple AS-PATHs for the replica connects on their own CPE devices to achieve the primary/replica relationship by making the AS-PATH lengths between multiple DC inconsistent.
- Static Routing:
- CLB (Active-active): Regardless of whether multiple lines are connected to the same DC point, as long as equal-cost routes are configured on the DC tunnel configuration of multiple lines, the traffic sent from the DC gateway on the cloud to the local data center under the cloud will be CLB;
- Primary/Replica (Active-standby): It supports the creation of redundant connections and active-standby tunnels under the same connect point.
Where is the interconnect IP actually configured?
The interconnect IP is configured at both ends of connection: the connect device of the DC point and the user's CPE device.
How does the cloud platform isolate different tenants to ensure data security?
The cloud platform private network provides users with an Internet network that uses MPLS_VPN technology. The DC point is equivalent to the PE node of MPLS. The user's DCion to cloud services supports VPC-level VRF technology isolation to ensure the user's data security.
What means are used to trigger the switching of redundant lines?
- Static routes are converged when the bidirectional BFD is configured.
- BGP routing triggers the switching through BGP's own convergence mechanism or when BFD is configured. For the configuration of BFD, you need to submit your request to the DC delivery engineer for backend configuration.
IPSLA cannot be configured. Reason: The connect point device has many connect lines. Configuring IPSLA will consume the performance of the connect device and may affect the normal use of all customers of the connect point.
Does the cloud platform support link aggregation control protocol (LACP)?
Not support.
Why do DC stress test results not meet expectations?
Reconfirm the created DC bandwidth.
Test the interconnect IP ping to see if there is packet loss. If there is severe packet loss, contact the cloud after-sales team to troubleshoot together.
You can check whether the local DC router is in full duplex mode.
You can use tools such as iPerf to try to initiate multiple processes on multiple clients and cloud servers to test whether the traffic meets expectations and eliminate application problems. If there is no progress in the troubleshooting, contact the cloud after-sales team for assistance.
What should I do if the DC to the cloud platform is interrupted?
- Confirm the interrupt mode:
The interconnect IP addresses cannot be pinged each other, and the MAC address of the other end cannot be received. If the DC has exception, the cloud platform after-sales service team receives an automatic alarm message. - Connect Interruption confirmed:
- If the DC owner is a cloud platform, report the problem to the cloud platform after-sales team as soon as possible. Generally, after an automatic alarm is received, the DC fault after-sales team reports the problem to the operator immediately.
- If the DC owner is a user, the customer needs to report the problem to the operator. If cooperation from the cloud platform is required, contact the after-sales team.
What should I do if the specific IP range of the DC cannot be pinged?
- Confirm whether the network is a newly enabled IP range. If yes, the cloud needs to configure the route.
- You can check whether the route table on the cloud has an inaccessible IP range pointing to the DC gateway and confirm whether the client machine has a route pointing to the DC.
- You can confirm whether the cloud has associated security groups, network ACLs and other security features, and confirm whether the client machine has set security features such as iptables.
- If no exceptions are found during the troubleshooting, provide the troubleshooting information to the after-sales team for follow-up.
What information do I need to provide when I contact customer service to solve a problem?
If you cannot solve the problem yourself, contact us and provide the following information to help improve the troubleshooting efficiency:
- Description of the problem and whether it can be reproduced.
- The source/destination IP pair in question.
- Regional information for DC.
- DC gateway ID.
- Ping/traceroute screenshot during failure.
- Information about iptables, NAT, etc.
- Topology diagram is connected.