Detailed operational procedures for Alibaba Cloud ECS, covering login, resource management, migration, upgrades, and troubleshooting.
The guide explains how to access the Apsara Stack management console, manage ECS resources, and perform routine tasks such as creating, starting, stopping, and migrating virtual machines. It also describes disk and snapshot operations, security group configuration, and the Sea Chart system for performance analysis.
Further sections cover hot migration of VMs and cloud disks, upgrade constraints for GPU and FPGA clusters, instance local‑disk repair, daily alarm handling, and comprehensive inspection procedures for cluster health and resource usage. This manual is written in Chinese.
| Product version | V3.10.0 |
|---|---|
| Document version | 20220106 |
| Minimum supported GPU cluster version | V3.3 |
| Minimum supported FPGA cluster version | V3.5 |
The default allowed downtime is 300 ms, with a configurable range of 200 ms to 2000 ms.
No. GPU clusters do not support hot upgrades; they require a full reboot of the NC servers before upgrading.