Service Level Agreement
Service Level Agreement
Last Revised: February 21, 2023
The Service is currently hosted on Amazon Web Services ("AWS"). The Service will be available 100.00% of the time that AWS and sub-vendors are functioning correctly.
There has been no Service maintenance downtime, ever. Should such downtime ever become necessary, Provider will use its best efforts to notify all account owners by email of the scheduled maintenance date and window and purpose.
If any User does not experience the promised level of availability in a single month, and User provides convincing proof that such failure is attributable to a fault in the Service and not to any other cause (and is therefore a "Service Failure"), Provider will grant User "Extra Credits", calculated over the (prorated) Credits for the month in which the Service Failure took place:
Availability of <100% to 99.9% → 5% Extra Credits;
Availability of <99.9% to 99.8% → 10% Extra Credits;
Availability of <99.8% to 99.7% → 15% Extra Credits;
Availability of <99.7% → 20% Extra Credits;
Provider will allow User to cancel the Service and receive a pro-rata refund of the fees paid, in response to any of the following: (a) more than 1 Service Failure during any 3 month period; (b) in case the Availability percentage of 99.6% is not met in a single calendar month.
UPTIME/AVAILABILITY
Lifetime Uptime: 99.9968%
Historical uptime:
| Year | Sovren planned downtime | Unplanned outages | Reason |
|---|---|---|---|
| 2023 | 0 minutes | 0 minutes | |
| 2022 | 0 minutes | 97 minutes | Intermittent 503 and 429 errors |
| 2021 | 0 minutes | 0 minutes | |
| 2020 | 0 minutes | 0 minutes | |
| 2019 | 0 minutes | 33 minutes | Intermittent timeouts occurred from customer abuse. See below for details. |
| 2018 | 0 minutes | 0 minutes | |
| 2017 | 0 minutes | 0 minutes | |
| 2016 | 0 minutes | 25 minutes | General intermittent AWS outage due to DDOS/DNS attack |
| 2015 | 0 minutes | 0 minutes | |
| 2014 | 0 minutes | 0 minutes | |
| 2013 | 0 minutes | 27 minutes | General intermittent AWS outage |
On 2019/09/27 at 10:01 AM EST, a customer sent the US SaaS system requests at more than 1,000 times the allowed throughput. This caused intermittent timeouts for some US based customers until the system was fully stabilized by 10:34 AM EST. Normally we are able to stop such problems before they overwhelm all of our servers, but this time we bungled our response, resulting in intermittent outages for approximately 33 minutes. As a result of this incident, Provider wrote an AI layer to continuously monitor and automatically protect the Service in realtime.
PERFORMANCE/STABILITY
Lifetime 100% Performance: 99.9999% (zone adjusted)
Periods of reduced performance:
| Year | Time | Product | Description |
|---|---|---|---|
| 2023 | 0 minutes | ||
| 2022 | 0 minutes | ||
| 2021 | 0 minutes | ||
| 2020 | 100 minutes | AI Matching | Intermittent slowness and timeouts in AI Matching due to a hardware failure in third party vendor's systems |
| 2019 | 75 minutes | AI Matching | Slowness and occasional errors in one availability zone in AI Matching due to third party vendor's systems |
| 2018 | 195 minutes | Conversion for parsing | Partial outage related only to conversion of some PDF documents in one availability zone; attributable to third party software (fixed) |
| 2017 | 0 minutes | ||
| 2016 | 0 minutes | ||
| 2015 | 0 minutes | ||
| 2014 | 0 minutes | ||
| 2013 | 0 minutes |
In the event that AWS changes, discontinues, deprecates or terminates its services in a way that adversely impacts Sovren’s ability to continue the Service, Sovren reserves the right to change or terminate the Service. Sovren will give Users as much notice as reasonably possible and, in the event of a termination of Service, will provide User a pro-rata refund of the lesser of the remaining unused time or the remaining unused Credits.