The following tables describe high-risk operations that need to be considered during the O&M stage of each component.
Cluster High-risk Operations
Action Name | Operation Risk | Risk Level | Risk Mitigation | Major Operation Observing Items |
Binding EIP | This operation exposes master nodes (such as the Doris FE) of the cluster service to the public network, increasing the risk of network attacks from the Internet. | ★★★★★ | Ensure that the bound EIP is a trusted public access IP. Verify if the corresponding ports have set security group rules, allowing only trusted IPs to access these ports. It is not recommended to allow 0.0.0.0 to access in the inbound direction rule. | None |
Opening cluster 22 port security group rule | This operation increases the risk of users exploiting port 22 for vulnerability attacks. | ★★★★★ | Set security group rules for the open port 22, allowing only trusted IPs to access this port. It is not recommended to allow 0.0.0.0 to access in the inbound direction rule. | None |
Deleting cluster or cluster data | This operation can lead to data loss. | ★★★★★ | Please confirm the necessity of this operation before deletion, and ensure that data backup has been completed. | None |
Scaling down cluster | This operation can lead to data loss. | ★★★★★ | Please confirm the necessity of this operation before scaling down, and ensure that data backup has been completed. | None |
Unmounting disk or formatting data disk | This operation can lead to data loss. | ★★★★★ | Please confirm the necessity of this operation before proceeding, and ensure that data backup has been completed. | None |
High-risk Operations in YI-MapReduce Manager
Action Name | Operation Risk | Risk Level | Risk Mitigation | Major Operation Observing Items |
Modifying log level | If modified to DEBUG, the running speed of the Manager will noticeably slow down. | ★★ | Confirm the necessity of the operation before modification, and revert to the default setting timely. | None |
Restarting underlying service with "Also restart upper-layer service" checked | This operation will interrupt the upper layer service business, affecting the management, maintenance, and business of the cluster. | ★★★★ | Confirm the necessity of the operation before execution, and ensure no other maintenance operations are executing concurrently. | Observe for any unresolved alarms, check if cluster management and maintenance are normal, and if the business is operating normally. |
Restarting service | The service will be interrupted during the restart. If "Also restart upper-layer service" is checked, any upper-layer service that depends on this service will be interrupted. | ★★★ | Confirm the necessity of the restart before execution. | Observe for any unresolved alarms, check if cluster management and maintenance are normal, and if the business is operating normally. |
Modifying default SSH port of node | Modifying the default port (22) will cause the Create Cluster, Add Service/Instance, Add Host, Reinstall Host functions to be unusable, and cause the cluster health check results, including node trust, to be inaccurate. | ★★★ | Change the SSH port back to the default value before performing related operations. | None |