How to handle planned or unanticipated events that affect the running status of your instance

This article introduces how to deal with planned or unexpected events that affect the running status of the instance, and focuses on the specific steps. The content of this article is very compact, and I hope you will study patiently.

System events

System events are planned or unanticipated events that affect the running status of the instance. Specifically refers to events such as restarting, stopping, or releasing ECS ​​instances due to detection of update maintenance, illegal operations, system failures, software and hardware failures, due to prepayment expiring or accounts in arrears.

Overview

Instances are the core basic components for building business applications. When you complete instance selection, purchase, initial configuration and start After the business is deployed, the health status of the instance plays a key role in your business continuity and system reliability. In order to improve system reliability, system performance and security protection capabilities, ECS will proactively perform routine maintenance on the physical server where the instance is hosted and repair potential system failures. When a physical server failure or hidden danger is detected, ECS will live-migrate the instance to a healthy server online to maintain the healthy running status of the instance. These maintenance are routine maintenance. Routine maintenance is different from system events. Routine maintenance is usually completed without any awareness. Maintenance notifications will not be sent and the instance will not be affected.

When a system event occurs, ECS will promptly send you notifications, response measures, event cycles and other information. For planned system events, information such as the impact of the event on the instance and the expected execution time will be informed in advance. You can back up data in time before executing system events, make application-level preparations, and prevent potential impacts on business reliability and continuity. For processed system events, you can query the historical system events of the instance in the past two months to obtain fault diagnosis and review analysis data.

Usage restrictions

Except for instance type families that have been discontinued, such as c1, c2, m1, m2, s1, s2, All instance type families, such as s3 and t1, support the system event function. For more details, see Instance Type Family.

Event types

System event types are shown in the following table.

Event status

In the life cycle of system events, there are the following event statuses.

Event window

System events have the following window period:

User operations Window period: The time period between the initiation of the system event plan and its execution. Usually the user operation window period for system maintenance-related events is 24 to 48 hours. The operation window period for stopping prepaid instances due to expiration is 7 to 15 days. The operation window period for stopping due to account arrears is usually less than 1 hour.

You can choose the recommended response method to repair the system event in advance based on the impact of the event on the business, or wait for the default response action to be performed. For system events triggered by ECS proactively repairing system failures, ECS will send you event notifications in advance according to the system maintenance operation plan.

Event execution window period: If you do not respond to system events in advance, repairing faulty system events will usually be completed automatically within 6 hours of entering the event execution window period. System event instances caused by billing issues will be immediately After stopping, classic network type instances will be released after 7 days and VPC type instances will be released after 15 days. You will then receive the execution results of the system event.

Note

For unexpected system events caused by non-technical factors such as sudden failures or illegal operations, there is usually no user operation window period. A short event execution window. You can query historical system events for fault diagnosis, cause analysis, or data recovery.

View system events

If there are planned system events, a prominent mark will appear on the pending event button on the ECS console to remind you to check .

Log in to the ECS management console.

In the left navigation bar, click Overview.

In the common operations area of ​​the overview page, click the pending event.

Click the system scheduled event page, and you can see related information such as instance ID, region, instance running status, event type, recommended actions, etc. You can perform the required operations in the operation column.

API operation: DescribeInstancesFullStatus queries the system events of the instance.

View historical system events

On the All Events page, you can view all system event records that occurred in the past two months, which are faults Diagnosis and review analysis provide data support.

Log in to the ECS management console.

In the left navigation bar, click Overview.

In the common operations area of ​​the overview page, click the pending event.

In the left navigation bar, click All Events. On the All Events page, click System Scheduled Events > Instances to view related information such as instance ID, event type, and event status.

API operation: DescribeInstanceHistoryEvents queries historical system events.

Subscribe to event notifications

All system events can be configured with alarm rules through cloud monitoring, and you will be notified in time when an event occurs. For more details, see the Cloud Monitoring document Cloud Product System Event Monitoring.

Response Suggestions

You can respond to system events through console operations or API calls. Managing ECS ​​based on system events can improve your underlying awareness of ECS and optimize ECS instance operation and maintenance.

The above is the detailed content of How to handle planned or unanticipated events that affect the running status of your instance. For more information, please follow other related articles on the PHP Chinese website!