"Services": are vCDN, vServer, vStorage, vLB, vDB, vMonitor and vContainer (K8S).

• 1.1."Client"/ "Service User": means any individual or organization using VNG Cloud's Services.
• 1.2.

  "Monthly Uptime" of vCDN, vServer, vStorage, vLB, vDB, vMonitor: means the period of time calculated by subtracting the number of minutes the Service System is in 'Unavailable State' (as defined below) due to the error of VNG Cloud and not including the time for maintaining, discontinuing Service due to Force Majeure events and other circumstances excluded according to the regulation in this SLA.

  "Monthly Uptime" of vContainer (K8S): the total time the service is used by all nodes during the billing cycle, minus the minutes when the vContainer Service is in the 'Unavailable State' (as defined below) due to VNG Cloud errors. It does not include maintenance periods, service interruptions due to Force Majeure Events, and other cases excluded as specified in this SLA.

• 1.3."Unavailable State"
  • (a) vServer Service Calculation:
    • - Downtime of the vServer. 

    • - Time when the VNG Cloud network is unable to connect to the ISP. 

      (Exception case: The time when the ISP network cannot connect to the end user is excluded from the calculation.)

  • (a)vContainer (K8S) Service Calculation:
    • - Downtime of the Node. 

    • - Time when the VNG Cloud network is unable to connect to the ISP. 

      (Exception case: The time when the ISP network cannot connect to the end user is excluded from the calculation.)

  • (a)VKS Service Calculation:

    • - Downtime of the Control Plane managed by VNG Cloud, defined as the total time (in minutes) during which a cluster on VKS cannot connect to the Kubernetes API Server. 

      (Exception case: Downtime may occur due to force-upgrades of end-of-life VKS versions, required-upgrades, or auto-upgrades of Kubernetes versions.) 

    • - Time when the VNG Cloud network is unable to connect to the ISP. 

      (Exception case: The time when the ISP network cannot connect to the end user is excluded from the calculation.)

• 1.4."Error Rate Storage": is the percentage of error calculated by the total number of failed requests (requests) divided by the total number of requests in 5 minutes (block of 5 minutes); The total number of blocks in the payment cycle is calculated by the total number of time using the Services in minutes divided by 5 and not exceeding 8,928 blocks (equivalent to 31 days / month). The error times counted when the request returned was a value of "InternalError" or "ServiceUnavailable".
• 1.5."Error Rate CDN": is the rate of arising errors measured by the CDN system through the error code HTTP Code Status 500 (Internal Server Error) is recorded and reported on the Portal system on the total number of requests arising in the billing cycle.
• 1.6."Monthly Uptime Rate": calculated as 100% minus the "Error Rate" in one (1) billing cycle, or calculated as 100% minus the "Monthly Downtime Rate".
• 1.7."Monthly Downtime Rate" is calculated by dividing the total time the system is in the Unavailable State by the total time in the month, then multiplied by 100.
• 1.8."Force Majeure Event": means an event that happens in an objective, unpredictable and irreparable manner despite the application of all necessary measures and in the permissible capacity, including but not limited to natural disasters, enemy sabotage, extreme weather phenomena, sabotage actions, protests, strikes, and bans from Government Authorities.
• 1.9."Error Rate vMonitor": average downtime of all alarms on a resource or log project.
• 1.10."vMonitor Downtime": is calculated by the time when the alarm is not working right on each resource or log project and then division for a total a real Alarm minutes existed in month and then multiplication for 100%.
• 1.11.The time when Alarm not working right: is calculated from the Alarm session failed when transferring notification to the session when Alarm working.
• 1.12.VNG Cloud provides two SLA options for vServer services, including (1) Multi-Region Level SLA (applies to compute instances deployed across multiple regions) and (2) Single Region SLA (applies to single compute instances deployed within a single region). 
  • (a)Single Region: A deployment model in which resources, applications, or services are confined to a single geographic area (region). A region typically consists of one or more data centers interconnected through a low-latency network. Each region operates independently from other regions, ensuring infrastructure isolation, including geographical location, power supply, water systems, and other foundational elements. 
  • (b)Multi-Region: A deployment model in which resources, applications, or services are distributed across multiple geographic regions, starting from two or more regions. Each region is an independent data center, connected through a low-latency, high-availability network with fault-tolerant capabilities.

• 3.1.

  Service Commitment

  Service Commitment	Rate	Unit	Equivalent
  vCDN	99,90%	CDN Domain	Measure the number of failed requests on the total requests.
  vServer/vDB (Single Region)	99,94%	1 Virtual host	Equivalent to 26 minutes 05 seconds
  vServer/vDB (Multi-region)	99,99%	1 Virtual host	Equivalent to 4 minutes 21 seconds
  vLB	99,96%	1 Load Balancer	Equivalent to 17 minutes 23 seconds
  vContainer (K8S)	99,94%	Cluster	Depends on the actual usage time in a month.
  vStorage	99,99%	Project	Equivalent to 0 Block fail 100% or 1 Block fail 80%
  vMonitor	99,90%	Resource/Log Project	43 mins and 28 seconds
  VKS	99,95%	Cluster	21 mins and 44 seconds

• 3.2.

  Compenstation for Commitment Violations

  If the Service fails to meet the Monthly Uptime Rate commitment, as stated above, VNG Cloud is responsible for providing the corresponding Service to the Customer. The compensation for customers is calculated as a percentage of the value during the billing period of that month.

  - vServer/vDB: For Single Region service

  Monthly Uptime Rate	Downtime Equivalent	Service compensation rate / 1 virtual server
  From 99,90% to less than 99,94%	From 26 minutes 05 seconds to 43 minutes 28 seconds	10%
  From 95% to less than 99,90%	From 43 minutes 28 seconds to 01 day 12 hours 13 minutes 27 seconds	30%
  Below 95%	More than 01 day 12 hours 13 minutes 27 seconds	100%

  - vServer/vDB: For Multi-Region services, if both regions experience downtime simultaneously, the service compensation rate per virtual server is as follows:

  Monthly Uptime Rate	Downtime Equivalent	Service compensation rate / 1 virtual server
  From 99,90% to less than 99,99%	From 43 mins 49 seconds to 4 mins 21 seconds	10%
  From 95% to less than 99,90%	From 43 mins 28 seconds to 01 day 12 hours 13 mins 27 seconds	30%
  Below 95%	More than 01 day 12 hours 13 minutes 27 seconds	100%

  - vLB:

  Monthly Uptime Rate	Downtime Equivalent	Service compensation rate / 1 vLB
  From 99,90% to less than 99,96%	From 17 minutes 23 seconds to 43 minutes 28 seconds	10%
  From 95% to less than 99,90%	From 43 minutes 28 seconds to 01 day 12 hours 13 minutes 27 seconds	30%
  Below 95%	More than 01 day 12 hours 13 minutes 27 seconds	100%

  - vContainer (K8S):

  Monthly Uptime Rate	Downtime Equivalent	Service compensation rate / 1 virtual server
  From 99.90% to less than 99.94%	vContainer (K8S) downtime depends on actual monthly usage	10%
  From 95% to less than 99,90%		30%
  Below 95%		100%

  - vCDN:

  Monthly Uptime Rate	Service compensation rate / Domain CDN
  From 99% to less than 99,90%	5%
  From 95% to less than 99%	10%
  From 90% to less than 95%	30%
  Below 90%	100%

  - vStorage:

  Monthly Uptime Rate	Service compensation rate / Project
  From 99% to less than 99,99%	10%
  From 95% to less than 99%	30%
  Below 95%	100%

   

  - vMonitor:

  Monthly Uptime Rate	Service compensation rate / 01 Resource or Project Log corresponding problem
  From 99,00% to less than 99,90%	10%
  From 95% to less than 99,00%	25%
  Below 95%	100%

  - VKS:

  Monthly Uptime Rate	Service compensation rate / 01 Cluster	Compensation
  From 99,00% to less than 99,95%	10%	Calculated based on the downtime value of the control plane
  From 95% to less than 99,00%	25%	Calculated based on the downtime value of the control plane
  Below 95%	100%	Calculated based on the downtime value of the control plane

• 4.1.The Service is inaccessible due to the fault of Clients, users and/or third party such as: expired domain name, Domain Name Server error (DNS), misconfiguration, operating system error, computer virus, spyware, unavailable Client’s network service, iptables system, firewall, etc.
• 4.2.The Service is blocked because the content(s), file(s) Client stores on VNG Cloud’s system violate Vietnamese law, is accused of intellectual property infringement and/or other VNG Cloud’s regulations; Client’s use of source code affects VNG Cloud’s system; Client does not make payment for the service fee for VNG Cloud.
• 4.3.Illegal access to other services which are not managed by Client.
• 4.4.The Service is inaccessible because Client’s use of the service is not in accordance with the standard recommended by VNG Cloud.
• 4.5.Due to Client’s request to stop the system.
• 4.6.The Service is inaccessible due to errors from Client’s devices.
• 4.7.The Service is inaccessible due to Force Majeure Event.
• 4.8.When Client uses the sources/services of VNG Cloud to attack other systems and is detected, VNG Cloud will notify to stop the service (for example: DDOS, intentional attack, virus distribution).
• 4.9.Service is suspended because the Client does not pay the subscription fee.
• 4.10.Service is interrupted due to the maintenance, which has been notified to the Client in advance.

• 5.1.System maintenance is mandatory to ensure the continuity of the Service provided to Client. The maintenance can be scheduled on any day of the week (including weekend) and at any time of the day. However, VNG Cloud shall endeavor to carry out the maintenance at the time that least affects Client’s use of the Service.
• 5.2.VNG Cloud shall notify Client in advance via email on VNG Cloud system or on the homepage https://vngcloud.vn/
  • (a)Three days in advance before expected day of maintenance.
  • (b)In emergency maintenance cases, VNG Cloud shall endeavor to notify Client 30 minutes in advance via email or telephone.
  • (c)For some incidents beyond the control, VNG Cloud shall notify Client within 4 working hours since the time the incident occurs.
• 5.3.VNG Cloud commits the total amount of maintenance time affecting the Service of Client in one (1) month does not exceed two (2) hours.

• 6.1.Forms of contact for receiving and solving problems:
  • (a)Via 24/7 support call center: 1900 1549
  • (b)Via website or email: https://vngcloud.vn; support@vngcloud.vn
• 6.2.Response time:
  • (a)Response time for troubleshooting request: within 15 minutes since the time the request is received.
  • (b)Response time and troubleshooting time for the Client’s problems: within 15 minutes up to 48 working hours since the time of responding to troubleshooting request.
  • (c)For serious problems, response time can be longer, but not exceed 72 working hours since the time of responding to troubleshooting request.

Forms of contact for receiving complaints: in document or send email via support@vngcloud.vn

• 7.1.Content of complaints: in order to save the Client’s time, when sending complaints to VNG Cloud, the Client should sufficiently provide the following information:
  • (a)Email or Document subject must be clearly stated: "Request to compensate according to SLA commitment";
  • (b)The information of the account affected;
  • (c)Specific date and time and the moments when the system, and Service used by the Client fall into the Unavailable State;
  • (d)Logs recording errors or screenshots at the time the error occurs (the Client’s confidential information should be blurred or hidden);
  • (e)Damages to the Client and the value of damages accompanied by relevant documents and invoices;
  • (f)The level of compensation that the Client requires.
• 7.2.Time to receive complaints: within three (3) months since the incident occurs; more than the above-mentioned time, VNG Cloud shall not accept and resolve any relevant complaints.
• 7.3.Time to respond and resolve complaints: within fifteen (15) days since the day the complaint is received.
• 7.4.In case the Client is dissatisfied with the resolution documents of VNG Cloud, the Client can resend feedback to VNG Cloud; in this case, the Client’s feedback shall be received and processed as a new Complaint.