Service Level Agreement Tweakwise - Tweakwise

Service Level Agreement Tweakwise

1. Service Level Agreement

1.1 Operations and technical support: Online support portal, onboarding, email and phone support during Normal Business Hours - 09:00 to 17:00 Central European Time (CET), during each Business Day.

1.2 Optimizers guarantees a minimum uptime of 99.9% for the frontend API.

1.3 The following maintenance windows apply:

  • Acceptance platforms and services: every 3rd Tuesday of the month between 4.00 (CET) and 6.00 (CET)
  • Production platforms and services: every 4th Tuesday of the month between 4.00 (CET) and 6.00 (CET)

1.4 We reserve the right to deviate from the maintenance windows mentioned above in case of:

  • Critical updates
  • Other issues that impact the confidentiality, integrity and availability of our platforms and services.
  • Other planned maintenance

1.5 Status Monitoring: monitoring the platform and services availability.

1.6 Service Desk response time:

  • Normal priority: best effort during Normal Business Hours. Normal priority refers to items 3, 4 and 5 of below prioritization matrix.
  • High priority: < 8 hours during Normal Business Hours. High priority refers to items 1 and 2 of below prioritization matrix.

1.7 Prioritization matrix:

  • Fatal; Any issue that affects all users, preventing business-critical day-to-day activities
  • Critical; Any issue that affects most users, greatly limiting day-to-day business activity.
  • Disruptive; Any issue that affects a single user and/or impacts the performance of a critical business activity.
  • Other; Any issue not related to a critical business activity

We assume ‘normal’ use of our software and connections within the designed specifications and limits.

To best help all our customers, we ask you to use the access to our (operational and technical) support services, professional services and resource consumption fairly.

Please refer to below support procedure:

  • Please visit support.tweakwise.com to submit any questions or chat with one of our qualified employees.
  • Please call in case of high priority matters (as indicated above) the following number +31883035730
  • In case of security incidents please send an email to: [email protected]


2. RESPONSE TIME API CALLS

2.1 Optimizers strives to guarantee the following Tweakwise API response time between 07:00 and 22:00 (CET) daily:

  • Percentile: 99.9%
    Response time: 2000
    99.9% of requests result in a response within 2000 ms
  • Percentile: 99.5%
    Response time: 1500
    99.5% of requests result in a response within 1500 ms
  • Percentile: 99%
    Response time: 1000
    99% of requests result in a response within 1000 ms
  • Percentile: 95%
    Response time: 500
    95% of requests result in a response within 500 ms
  • Percentile: 50%
    Response time: 250
    50% of requests result in a response within 250 ms

2.2 Regarding the response times the following demarcation apply:

  • Agreed parameters as reflected in the agreement with a maximum of:
    • 100,000 products
    • 2,250 categories
    • 1,000,000 requests per day
  • Guarantees apply to instances where no beta features are running
  • The default (max) number of 8 properties per product in response
  • Measured percentiles over a month.
  • Measured performance excluding maintenance window and disturbances (patches, updates, failures)
  • With throttling we limit the maximum load that an instance can generate at any given time. With throttling we do not mean a certain number of requests per second. Because that implies nothing about the load on an instance. It is also about the severity of a request.
    With throttling we limit the number of concurrent requests an instance can have at any given time. We call these workers:
    • One worker can process exactly one request at a time. For example: In one second, a worker can handle 10 requests if they last 0.1 sec each.
    • If all workers of the Client are occupied a request is placed in the queue.
    • If a request is in the queue for more than 10 seconds it will be fired with the code: 429 TOO MANY REQUESTS (https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/429)


OTHER NOTES

3.1 The throttling method can be compared to a shop counter/checkout. Only one person can always be helped at a counter, and not all requests for help take the same amount of time. When all counters are occupied, a queue arises. Optimizers can give a customer more capacity by reserving more counters.

3.2 The following fault reports (calls) can be made:

  • Tweakwise unreachable
  • Error message on Tweakwise environment

3.3 The following reports/issues are not covered by this service level agreement:

  • Adjustments to imports and exports
  • All matters relating to test environments