UHadoop Service Level Agreement for Managed Hadoop Cluster

This UHadoop Service Level Agreement for Managed Hadoop Cluster (hereinafter referred to as “this Agreement”) stipulates the general service level indicators and service attention items provided by UCloud Global to customers. This agreement is a supplement to the “UCloud Global User Agreement” and the “Cloud Service Legal Statement and Privacy Terms”. By stamping, clicking to confirm on the network page, or selecting to accept this agreement in other ways (such as clicking to confirm and agree to use the “UCloud Global User Agreement”), you have reached an agreement with the company and agree to accept all the contents of this agreement. If there is any inconsistency between the text of the party’s stamp and the terms of service that is selected to be accepted by clicking on the network page or in other ways, the text of the two parties’ stamp shall prevail. Before accepting this agreement, please read the full text of this agreement carefully. After reading, if you continue to use the UHadoop service for managed Hadoop clusters, it means that you accept this agreement and agree to be bound by it; if you disagree with this agreement, you can choose not to use the UHadoop service for managed Hadoop clusters.

1. Data Storage Durability

1.1 The data storage durability of the managed UHadoop cluster is 99.99%. If the service durability does not reach the promise, compensation will be paid in accordance with the service compensation clause, if the user modifies the number of copies of the data and it is less than the default number of copies, they will not enjoy this service durability promise.

1.2 Definition of Data Storage Durability: The probability of data not being lost in the storage state during the contract period. Durability of 99.99% means that for every 10,000 bytes of user data per month during the contract period, the probability of data not being lost per month during the contract period is 99.99%, i.e., only 1 byte of data may be lost each month. Based on the calendar month as the statistical cycle, less than a month is counted as a month.

2. Data Destructiveness

2.1 When you request data deletion or before discarding or reselling the device, UCloud Global will completely delete all user data and cannot restore it through advanced zeroing operations and perform demagnetization on scrapped hard drives.

2.2 The scenario of complete deletion of user data does not include the situation where UCloud Global retains user data for 7 days after obtaining UCloud Global‘s consent to prevent important data from being deleted due to various reasons for the user’s inability to renew the lease in time.

3. Data Privacy

3.1 UCloud Global has adopted effective isolation methods to ensure that user data in the same resource pool is not visible to each other.

3.2 The managed UHadoop cluster is isolated according to the account, and network isolation technology is used to ensure that nodes and data between different users are not visible and cannot be accessed through the internal network.

4. Data Informed Assent

4.1 You can see the name of the available area where the managed UHadoop cluster is located in the product control panel. You can choose the available area you need. For example, North China 1 Available Region B is located in Beijing.

4.2 All data centers should comply with local laws and the laws of the People’s Republic of China.

4.3 Except for the needs of government regulatory department supervision and audit or to cooperate with security forensic investigation, user data cannot be illegally provided to any third party without the customer’s consent. In order to ensure your safety in using this service and to continuously improve service quality, the company will record and retain related behavior logs of your login and usage of this service and will not present user personal information data.

5. Business Audibility

5.1 In accordance with national laws and regulations, in order to cooperate with government regulatory authorities’ regulatory inspections, compliance, or evidence investigations, etc., the company can provide the relevant information of your business running on cloud services, such as the operation logs of key components and the operation records of operation and maintenance personnel.

6. Business Function

6.1 The service features provided by the managed UHadoop cluster to users are as displayed on the console.ucloud-global.com website. For a complete introduction and operation instructions, see the product manual and help guide.

6.2 UCloud Global will promptly notify you via text message, email, or website, etc., if it changes the version or features of the service, so that you can make corresponding adjustments in time.

7. Business Availability

7.1 The company promises to you the service availability of the managed UHadoop Cluster is 99.95%. If the service availability does not reach the promise, compensation will be paid in accordance with the service compensation clause. If the user operates non-standardly on the managed UHadoop cluster, the component services of non-standard operation will not promise the above service availability commitment.

7.2 Definition of Service Availability: The service availability is the probability of the customer’s monthly cloud service business available time during the contract period, that is, the actual available time per month / (actual available time + unavailable time) per month. The definition of unavailable time is from the time when the user cannot use the cloud service until the cloud service is restored to normal. Based on the calendar month as the statistical cycle, less than a month is counted as a month, and it is measured in minutes. Availability of 99.95% means that the single user’s business available time each month should be at least 30 days * 24 hours * 60 minutes * 99.95% = 43178.4 minutes, that is, there is a maximum of 30 days * 24 hours * 60 minutes * 100% - 43178.4 = 21.6 minutes of unavailable time each month. The statistical unit of cloud service business unavailability is the component service on the managed UHadoop cluster. If the unavailable time of the component service reaches 1 minute or more, it is counted as unavailable and counted into the unavailable time. If the unavailable time is less than 1 minute, it is not counted into the unavailable time.

8. Fault Recovery Ability

8.1 Through a series of fault control systems such as fault monitoring, fast positioning, automatic recovery, and notification, it ensures the fault recovery ability of cloud services. The means of fault recovery include: automatic disk replacement for bad disks, machine failure, automatic recovery of component processes that hang up, node migration.

8.2 The user should be informed to migrate when there is a risk of a fault in the running environment of the managed UHadoop cluster to avoid affecting the user’s business.

8.3 When the unavailability of cloud services is caused by statistical or billing logic, the statistical billing can be temporarily taken offline to prioritize the customer’s business. The system architecture design can support degradation to ensure business operations.

9. UCloud Global Disclaimer

For the following reasons causing the service to be unavailable, UCloud Global does not assume any responsibility:

9.1 Caused by system maintenance notified to the customer in advance by UCloud Global, including cutover, repair, upgrade, and simulated fault drill;

9.2 Caused by network, equipment failures or configuration adjustments outside any devices belonging to UCloud Global;

9.3 Caused by the customer’s application being hacked;

9.4 Caused by the customer’s improper maintenance or poor confidentiality resulting in data, password, password, etc. being lost or leaked;

9.5 Caused by negligence or operation authorized by the customer;

9.6 Caused by the customer not following the UCloud Global product documentation or use recommendations;

9.7 Caused by the customer violating the “UCloud Global User Agreement” leading to service suspension or termination, including service suspension or release due to arrears of the managed Hadoop cluster UHadoop, etc.

9.8 Faults caused by user usage issues, including but not limited to: improper configuration parameters, unreasonable resource usage, business logic bugs, etc.

9.9 User faults caused by Bugs in the open-source components of the Hadoop ecosystem;

9.10 Other force majeure factors. Force majeure and accidents refer to unforeseeable, unavoidable, and insurmountable objective conditions, natural and social phenomena, due to normal maintenance of the information network, connection failures of the information network, failure of computers, communications or other systems, power failures, war, natural disasters, government actions, etc.