Cost monitoring and alerting help identify cost issues and cost optimization opportunities. Therefore, enterprises should effectively monitor cost data and resource data through various monitoring tools, and notify relevant operations personnel or financial personnel through alerts. This helps to track and handle identified issues or optimization opportunities, avoid unexpected costs, optimize resource utilization, and achieve cost savings.
Budget Management and Alerting
Budget management for cloud costs is an effective way to control cloud costs. By budgeting, alerting, forecasting, rolling adjustment, budget-to-actual analysis, and optimization, a budget-based cost closed-loop management chain can be established.
Enterprises can use the Budget Management tool for pre-cloud cost planning and budgeting, monitoring and alerting during the process, and actual-to-budget comparison and analysis afterwards. For example, enterprises can:
Use multidimensional budget management for costs based on financial units, products, accounts, etc.
Conduct budget tracking for the consumption amount of cloud resources or the usage rate/coverage rate of resources to be deducted, based on cost budgets, usage rate/coverage rate budgets, and on-demand budgets.
Use cost forecast tools that support 12 months to provide forecasting suggestions for budgeting.
Set alerts for actual and forecasted costs respectively.
Set multiple alerts and send multiple alert notifications based on different thresholds.
Detect differences between actual and forecasted costs in a timely manner through a budget-to-actual comparison.
Enable Anomaly Detection and Alerting
Excessive consumption due to unexpected costs is a common cost control issue. It is necessary to periodically check whether consumptions are based on actual business needs. Enterprises can use the Anomaly Detection feature provided by Alibaba Cloud to promptly detect cost anomalies and quickly locate the causes of problems with the help of accompanying root cause reference information.
Anomaly detection is implemented using artificial intelligence algorithms. Enterprises can customize the sensitivity of anomaly detection and the system intelligently identifies cost fluctuation anomalies. This enables automatic monitoring, detection, and alerting of unexpected costs.
Monitoring and Alerting Available Credit Limit
The available credit limit refers to the amount of cash remaining in the user account, from which frozen or future deductions are subtracted. If a user is a credit control user, the available credit limit also includes the credit limit of credit control. It is recommended for enterprises to enable Balance Alarm to obtain timely information about the available credit limit.
Monitoring and Alerting High Bills
Enterprises should monitor high bills and promptly analyze the reasons for excessive expenses. The notifications tool can be used to alert when the daily bill of a postpaid product exceeds the alert threshold (calculated based on the daily bill fee as of 24:00 yesterday).
Monitoring Resource Plan Usage
Enterprises should set usage limits for deducted resources. The system will notify users based on the limit set by the user to avoid insufficient deducted resources. For monthly and dynamic monthly resources and total decremental resources, the alert rules are different:
For monthly and dynamic monthly resources, alerts are triggered once per period based on the threshold and once when the period is exhausted.
For total decremental resources, an alert is triggered once per survival period based on the threshold.
Monitoring Resource Usage
Monitoring the resource utilization of cloud resources helps grasp the level of resource consumption and enables timely tracking of scenarios such as resource overload, low resource utilization, and resource idle. This helps enterprises better manage the use of cloud resources. Based on monitoring, enterprises can analyze historical data to effectively predict changes in cloud platform capacity and provide early warnings.
Daily Monitoring of Cloud Resource Usage
It is necessary to establish daily monitoring of the resource utilization of cloud resources. Cloud users can use the cloud product monitoring capabilities of Alibaba Cloud CloudMonitor to observe the real-time or near real-time usage of cloud resources, and also set alarm rules.
High resource load on cloud resources usually means that elastic scaling of cloud resources may be required, Enterprises need to analyze resource high load situations to see if there is improper resource utilization. If there is improper resource consumption, architecture optimization or code optimization is needed. If resource consumption is normal, new resource approval can be carried out in accordance with the plan and standards established during the cost planning phase after financial approval.
Low resource load on cloud resources usually means low-load or idle resources, resulting in resource waste. After analysis, resources can be scaled down to avoid impacting business operations. Scaling down operations should be carried out during low-peak periods.
Monitor the resource change time for resources with elastic scaling rules to avoid excessive scaling or failure to scale in a timely manner due to improper rule settings and additional costs.
Unused computing resources that have been idle for a long time should be shut down and released. Similarly, unnecessary storage files, such as snapshots, should be deleted in a timely manner to avoid continuous resource waste.
Evaluate potential optimization opportunities for resource utilization based on continuous monitoring data. By analyzing monitoring data, traffic characteristics during peak and off-peak hours can be observed, guiding enterprises to use a flexible combination of various billing methods to save costs.
Use Enterprise Cloud Monitoring for Complex Deployment and Usage Scenarios
Enterprises can easily achieve unified monitoring of resources across Alibaba Cloud accounts using the resource directory. For hybrid cloud scenarios, it is recommended to use enterprise-level cloud monitoring, which is based on cloud resources, to efficiently and cost-effectively aggregate monitoring data for resources outside the cloud. This enables the rapid construction of an integrated monitoring system for resources both inside and outside the cloud. The Intelligent Water Level Analysis Report tool makes it easy for enterprises to obtain performance reports based on Alibaba Cloud big data and use artificial intelligence algorithms for in-depth analysis of product usage status.
Analyze Correlations Between Budget and Resource Monitoring Data
Budget monitoring data helps observe fund consumption and correlate it with specific usage of cloud products. By combining resource monitoring data, enterprises can analyze the timing and events of resource consumption and take appropriate measures. After setting up the budget plan, it is necessary to track the usage and provisioning of cloud resources during the budget period. If the number or consumption of cloud resources is lower than the budget plan, it may be due to unexpected issues during system launch or project stagnation, among other reasons. In such cases, the budget plan needs to be modified based on the actual situation.