Service Level Agreement

The following terms have specific meanings in this SLA. Measurements are taken by Dollu’s network monitoring, CDR and ticketing systems, which are the authoritative record for the purpose of calculating availability and service credits, unless a manifest error is shown.

• Availability: the percentage of minutes in a Service Month during which the service is not in Outage, calculated as total minutes minus Outage minutes, divided by total minutes.
• Outage: a period during which the service is unavailable or materially degraded because of a fault in Dollu’s network or platform, starting when the fault is detected by our monitoring or a P1 ticket is logged, whichever is earlier, and ending when service is restored.
• Service Month: a calendar month in which the service is in production for the full month.
• MRC: the monthly recurring charge for the affected service; for usage-only services such as voice termination and SMS, MRC means the average monthly usage charge for that service over the previous three months.
• Time to Mitigate (TTM): for DDoS protection, the time from attack detection to the start of mitigation that reduces attack traffic reaching the protected prefix.
• Response Time: the time from a ticket being logged with Dollu to a qualified engineer acknowledging it and beginning work.
• Planned Maintenance: work notified in advance in accordance with this SLA, which is excluded from Outage.

Each service family carries an availability target measured monthly at the boundary of Dollu’s network. Targets apply to Dollu’s platforms and interconnect edges; they do not extend to the public internet, your own equipment or the networks of destination operators, which are outside our control and are addressed through routing rather than guaranteed.

• Core voice switching and wholesale termination platform: 99.99% availability.
• SIP trunking, including registration, signalling and media relay at our session border controllers: 99.99% availability.
• Numbers platform, covering inbound routing for DIDs, toll-free and cloud calling numbers, and the porting and provisioning portal: 99.95% availability.
• CPaaS APIs for voice, SMS, verify, lookup and video, including authentication and webhook delivery: 99.99% availability.
• Messaging platform, covering A2P submission and delivery-receipt handling at our SMSC: 99.95% availability.
• Managed circuits (MPLS, DIA, Ethernet private line): 99.9% for single-homed access, 99.95% for dual-homed access to one PoP, and 99.99% for diversely routed access to two PoPs.
• DDoS protection: Time to Mitigate of less than 60 seconds for volumetric attacks detected by our always-on scrubbing.
• Cloud hosting and unified-communications platforms: 99.95% availability unless a higher tier is purchased.

Availability is measured from Dollu’s monitoring probes, SBC and switch logs, and API gateway telemetry, sampled at intervals of one minute or less. For voice services, an Outage exists when call attempts to a working destination fail at our platform for more than 5% of attempts over a five-minute window, or when signalling to a customer trunk is lost. For APIs, an Outage exists when the error rate for valid requests exceeds 5% or median latency exceeds 2 seconds over a five-minute window.

Partial outages affecting a subset of routes, numbers or regions are pro-rated to the proportion of the customer’s service affected. Degradation caused by a single destination operator, or by an upstream carrier failure that our routing engine works around within the target, is not counted as an Outage.

Our NOC operates 24×7 from Noida with follow-the-sun coverage in London and Singapore. Tickets can be raised by portal, email to [email protected] or telephone. You assign a priority when logging a ticket and we may reclassify it, with an explanation, if it does not meet the definition. Response times apply around the clock for P1 and P2, and during business hours in your contracted region for P3 and P4.

Every customer has a named account manager and an escalation path to the NOC shift lead, the Head of Network Operations and, for P1 incidents exceeding 4 hours, the Chief Technology Officer. A root-cause analysis is provided within 5 business days of the closure of any P1 incident on request.

• P1, critical: total loss of service or a fault affecting a majority of traffic with no workaround. Response within 15 minutes, updates every 30 minutes, target restoration within 4 hours.
• P2, major: significant degradation such as failure of a major destination group, one-way audio across a trunk, or API errors affecting a subset of requests. Response within 1 hour, updates every 2 hours, target restoration within 8 hours.
• P3, minor: a fault affecting individual numbers, routes or features with a workaround available. Response within 4 hours, target resolution within 3 business days.
• P4, request: provisioning changes, questions, reporting requests and non-urgent issues. Response by the next business day, target completion within 5 business days.

Standard maintenance windows are Tuesday to Thursday between 22:00 and 04:00 local time of the affected PoP, and Sunday 00:00 to 06:00 for global platform changes. We give at least 5 business days’ notice of planned maintenance by email and portal, and 24 hours’ notice for emergency maintenance needed to fix a security or stability issue. Our platforms are engineered so that most maintenance is hitless; where a service impact is expected we will state its nature and estimated duration in the notice.

Planned and emergency maintenance notified in accordance with this section is excluded from Outage calculations. Maintenance that overruns its notified window, or that is carried out without notice, counts as an Outage from the end of the window or from the start of the impact, whichever applies.

If we miss the availability target for a service in a Service Month you may claim a credit calculated as a percentage of that service’s MRC. Credits are tiered by cumulative Outage duration beyond the target allowance: up to 4 hours over, 5% of MRC; between 4 and 8 hours, 10%; between 8 and 24 hours, 15%; between 24 and 72 hours, 20%; and beyond 72 hours, 25%. For DDoS protection, each attack for which TTM exceeds 60 seconds earns a 5% credit, capped at 25% per month.

For managed circuits with a stated repair objective, an additional 5% credit applies for each full 4-hour period by which restoration exceeds the objective, within the same 25% monthly cap. Response-time credits of 2% of MRC apply where a P1 or P2 ticket is not acknowledged within the committed time. Total credits for any service in any month are capped at 25% of that service’s MRC, are applied against future invoices, and are your sole and exclusive remedy for a service level failure.

This SLA does not apply, and Outage minutes are not counted, where the failure results from circumstances outside Dollu’s reasonable control or from matters that are your responsibility. The exclusions below are applied consistently, and we will tell you which exclusion applies if we reject a claim.

• Your equipment, software, configuration, local access circuits not supplied by Dollu, or the public internet between your site and our edge.
• Failure or filtering by a destination operator, national regulator, transit carrier or number-range holder, including operator-side spam blocking and DLT scrubbing.
• Suspension or blocking under our Terms of Service or Acceptable Use Policy, including fraud controls triggered by your traffic.
• Planned or emergency maintenance notified in accordance with this SLA.
• Force majeure events, including submarine cable cuts, power grid failure, natural disaster, government action and civil disturbance.
• Traffic in excess of the capacity, channels, TPS or API rate limits ordered, or traffic patterns materially different from your forecast.
• Services in trial, beta or free-of-charge use, and any service for which your account is in arrears.

To claim a credit, email [email protected] or raise a portal ticket within 30 days of the end of the Service Month in which the failure occurred, quoting the affected service, the ticket numbers logged during the incident and the outage period you believe applies. We will validate the claim against our monitoring and ticket records and respond within 15 business days. Approved credits appear on the next invoice. Claims submitted after 30 days, or for periods in which your account was suspended for non-payment, are not eligible.

If the same service misses its availability target in any three consecutive Service Months, or in any four months within a rolling twelve-month period, you may terminate that service without early-termination charges by giving 30 days’ written notice within 30 days of the third or fourth failure. This right is in addition to service credits and does not affect other services on your account or any charges already incurred.

Incidents should be reported to [email protected] or by portal ticket at any time. Credit claims, SLA questions and requests for incident reports can be sent to [email protected] or by post to Dollu Ltd, 2403A, Iconic Corenthum, Sector 62, Noida 201301, India.