CloudLens for EBS support for reporting cloud disk events-Elastic Compute Service(ECS)-阿里云帮助中心

CloudLens for EBS reports cloud disk events with attributes, descriptions, and handling suggestions for each event type.

Event attributes

Event attribute	Description
EventName	The system event name.
ResourceId	The cloud disk ID associated with the event.
ResourceType	The resource type associated with the event. Example: Clouddisk.
Description	The event description.
EventType	The event type. Valid values: Notification: triggered based on usage; automatically closed after reporting. Alert: triggered based on usage; must be manually closed. System exception: triggered by an underlying issue that affects disk usage.
EventLevel	The event level. Valid values: Critical: Critical Warn: Warning Info: Notice
EventId	The event ID, unique per resource.
EventStatus	The event status. Valid values: Active Recovered
StartTime	The event start time.
EndTime	The event end time.
RecommendAction	The recommended action. Valid values: ModifyDiskSpec: change the disk specification. CreateSnapshot: create a snapshot for the disk. ResizeDisk: resize the disk. AdjustProvision: change the provisioned performance values of the disk. ModifyInstanceSpec: change the interface type of the instance to which the disk is attached.
RecommendParam	Parameter for processing the event. For example, a recommended provisioned value for Cost Optimization events.

Note

To query the attributes of a specific event, go to the cloud disk events page in the EBS console or call the DescribeEvents operation.

Event handling

The following table lists cloud disk events and suggested actions.

Event name	Description	Report frequency	Event type	Event level	Suggestion	Recovery behavior
Data Protection	No snapshots were created for a cloud disk for an extended period. Without recent snapshots, you cannot restore the disk in the event of ransomware, which may cause data loss.	Every morning	Alert	Warn	Create snapshot manually.	After the cloud disk recovers, the system reports a new event in the Recovered state the next morning. The Active event remains.
Cost Optimization	If the provisioned performance (such as 1,000 IOPS) exceeds the actual workload (such as 800 IOPS), you are charged for the higher level, wasting cost. If the provisioned performance (such as 1,000 IOPS) is lower than the actual workload (such as 1,200 IOPS), the disk frequently bursts to handle excess I/O, increasing total cost. Adjust the provisioned values to balance performance and cost. See ESSD AutoPL disk.	Once a week	Alert	Info	Modify the performance configuration of an ESSD AutoPL disk. Modify a disk category.	After the cloud disk recovers, no events of the same type are reported.
I/O Hangs on cloud Disks	An I/O hang occurred when excessively high read/write I/O latency on a disk caused the operating system to become unstable or experience downtime.	Real time	SystemException	Warn	Check disk performance metrics to identify the cause. See View the monitoring data of a disk. To learn how Alibaba Cloud Linux detects I/O hangs, see Detect I/O hangs of file systems and block layers.	After the disk recovers, the system reports a new event in the Recovered state. The Active events remain.
Specification Mismatch Between the Instance and Disks	The total disk specifications on an instance exceed the upper limit of the instance type. Disk performance may be limited by the instance type. For example, if an ECS instance has a maximum IOPS of 60,000 and a disk with 100,000 IOPS is attached, the disk performance is limited by the instance type. Cloud disk specifications: see Block Storage performance. Instance type: see Instance family overview.	12:00 to 15:00 every day	Alert	Warn	When total disk performance on an instance exceeds the instance type limit, issues such as slow data processing and high response latency may occur. Change the instance type based on your requirements. See Upgrade the instance types of subscription instances and Change pay-as-you-go instance type.	After the disk recovers, no events of the same type are reported.
Total IOPS of Disks Reached Maximum IOPS per Instance	The total IOPS of cloud disks attached to an instance reached the instance upper limit. For example, an ECS instance has a maximum IOPS of 60,000 and two attached disks with 10,000 and 51,000 IOPS respectively. The total real-time IOPS reaches the instance maximum. Disk IOPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum IOPS for an instance: see the disk baseline IOPS and burst IOPS descriptions in the Instance type specifications table of Instance family overview.	Within 5 minutes after the event is triggered. Latency is in minutes.	Notification	Warn		When the event is triggered, a Recovered event is also reported, regardless of whether the Active event is handled.
Total BPS of Disks Reached Maximum BPS per Instance	The total BPS of cloud disks on an instance reached the instance upper limit. For example, an ECS instance has a maximum BPS of 150 MB/s and two attached disks with 100 MB/s and 60 MB/s BPS respectively. The total real-time BPS (160 MB/s) exceeds the instance maximum. Disk BPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum BPS for an instance: see the disk baseline bandwidth and disk burst bandwidth descriptions in the Instance type specifications table of Instance family overview.		Notification	Warn
IOPS of a Disk Reached Maximum IOPS per Instance	The IOPS of a disk reached the instance upper limit. For example, an ECS instance has a maximum IOPS of 60,000 and one of its attached disks has 70,000 IOPS. The real-time IOPS exceeds the instance maximum. Disk IOPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum IOPS of an instance: see the disk baseline IOPS and burst IOPS descriptions in the Instance type specifications table of Instance family overview.		Notification	Warn
BPS of a Disk Reached Maximum BPS per Instance	The BPS of a disk reached the instance upper limit. For example, an ECS instance has a maximum BPS of 150 MB/s and one of its attached disks has 160 MB/s BPS. The real-time BPS exceeds the instance upper limit. Disk BPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum BPS of an instance: see the disk baseline bandwidth and disk burst bandwidth descriptions in the Instance type specifications table of Instance family overview.		Notification	Warn
IOPS of a Disk Reached Maximum IOPS per Disk	The IOPS of a disk reached the disk upper limit. Disk IOPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum IOPS of a disk: see Block Storage performance.		Notification	Warn	Issues such as slow data processing and high response latency may occur on the cloud disk. For performance metrics of different disk categories, see Block Storage performance. View the monitoring data of a disk. Reduce the read/write IOPS or modify the disk category for higher performance. Resize the disk to improve performance.
BPS of a Disk Reached Maximum BPS per Disk	The BPS of a disk reached the disk upper limit. Disk BPS: see the View the near real-time monitoring data of a disk section in "Analyze disks". Maximum BPS of a disk: see the Performance of cloud disks section in "Block storage performance".		Notification	Warn
Non-4K-Aligned Read/Write Operation	Read/write operations on the disk are not 4K-aligned, which may degrade disk I/O performance. Note If read/write operations in partitions are not 4K-aligned, the disk may perform a read-modify-write across two 4K sectors. A single read/write may then involve multiple I/O operations, degrading disk performance.	12:00 every day	Notification	Info	See Resolve high disk I/O load on a Linux instance.
Slow I/O Operations on Disks	The disk has slow I/O operations that take 1 second or longer to complete.	Real time	Notification	Warn	Slow I/O may affect disk-dependent applications, such as slow website loading. Check whether your business is affected promptly.
Disk I/O Burst	Disk I/O bursts occurred on the disk, which may generate performance burst fees. See ESSD AutoPL disk.	Every 1 hour	Notification	Info	Check whether burstable I/Os meet your business expectations.
Burst Performance Fee Cap on Disks	The disk has a performance burst and the total I/O bursts triggered burst performance fee cap rules.	Every 1 hour	Notification	Info	Although burst performance fee cap rules help reduce performance burst fees, check whether burst I/Os meet your actual requirements to prevent unexpected costs.