Infrastructure
We guarantee that your server hardware will be functioning 99.9% of
the time in a given month, excluding scheduled maintenance. This
includes being powered but does not include an outage on an
individual power supply on your servers if the server is still
powered and functioning through the other power supplies.
Infrastructure downtime exists when a particular server is shut down
due to power, heat, or hardware failure problems. We guarantee the
functioning of all server hardware components and will replace any
failed component at no cost. “Hardware” means the processor(s), RAM,
hard disk(s), motherboard, NIC card and other related hardware
included with the server. Hardware replacement will begin once we
identify the cause of the problem. Hardware replacement is guaranteed
to be completed within four hour of problem identification.
Guarantee
will credit your account the pro rata portion of the monthly charge
for the server affected that falls short of the 99.9% uptime.
GPU-Level Uptime
In addition to the Infrastructure and Network Uptime guarantees above,
guarantees that each GPU server delivered under the applicable Customer
Success Order will be Available for at least 98.9% of each calendar
month (“GPU Uptime”), measured per server. GPU Uptime is measured as
the average availability of all GPUs within a server during a calendar
month. Individual GPU failures within a server shall not constitute a
server-level Uptime breach unless the server’s average GPU Uptime falls
below the guaranteed threshold. A GPU is “Available” when it is powered,
responsive to 's standard monitoring systems (polled at intervals no
more frequent than every five (5) minutes), and capable of executing
workloads. Individual periods of GPU unavailability lasting less than
five (5) consecutive minutes shall not be counted toward GPU Uptime
calculations.
EXCLUSIONS: GPU server Uptime shall not include periods
of unavailability resulting from:
- Scheduled maintenance windows (provided Customer is given at least five (5) business days' prior written notice);
- Customer's or its authorized users' acts or omissions, including misconfiguration, unauthorized software, or workload errors;
- Force majeure events;
- Third-party network, power, or infrastructure failures beyond 's reasonable control;
- Customer's failure to provide required access or cooperation;
- Previews, beta services, or experimental features; or
- Any maintenance or repairs necessitated by Customer's breach of the Agreement.
Guarantee
If monthly GPU server Uptime falls below 98.9%, shall credit Customer
against the following month's recurring GPU Fees for the affected
servers on the following schedule: up to 8 hours due to a server or
component failure that is replaced and remedied or replaced with a
spares pool server, 0 downtime credit, longer than 8 hours a credit of
the actual downtime.
SPARES POOL: will maintain full server spares and make
available spares to client within 8 hours if individual server downtime
is not resolved. The spares pool shall constitute 3% of the amount of
servers in the cluster rounded up to the nearest whole number.
CREDIT EXCLUSIVITY: GPU server Uptime credits are
Customer's sole and exclusive remedy for any failure to meet the GPU
Uptime guarantee, and shall be Customer's sole and exclusive remedy for
any Service unavailability. In no event shall Customer be entitled to
any additional damages, refunds, offsets, or other relief arising from
or related to GPU Uptime shortfalls beyond the credits expressly set
forth herein.
Network uptime
We guarantee that our GPU network will be available 99.9% of the time
each month, excluding scheduled maintenance. The GPU network consists
of the portion of the network that extends from the outbound port on
your GPU server to the outbound port of the data center border router
where it interconnects with the upstream ISP transit provider’s
equipment.
Guarantee
We will credit your account the pro rata portion of the monthly charge
for network downtime that falls short of the 99.9 % uptime. DDOS and
performance issues such as partial packet loss that does not result in
an outage are excluded.
Exclusions
- Scheduled maintenance on all equipment and operating system updates are excluded and not calculated in downtime.
- Support or issues due to applications outside of the hardware.
- Internet issues that are outside of the network (at the point of egress to the upstream transit provider).
- None of the guarantees in this SLA may be combined. For example, a hardware outage that also disrupts network activity will not qualify You for two (2) separate credits.
Credit requests
You must initiate a support ticket and request that a credit be
applied to your account. Merely initiating a support ticket related to
an outage will not result in any credit to your account. Initiating a
support ticket and requesting a credit is what triggers an event for
any potential credits; however, must still determine, in its sole
discretion, whether or not an eligible outage has occurred. If
determines that an eligible outage has occurred, then the ticket you
generated will be used to generate a credit.
Notice and Cure
Customer must notify in writing within five (5) business days after the
end of the calendar month in which the alleged Uptime shortfall
occurred, specifying the affected server and the nature of the claimed
shortfall. shall have ten (10) business days following receipt of such
notice to investigate and, if applicable, issue the credit. The customer
must open a ticket notifying of downtime when the downtime occurs and
that is when the downtime clock starts for credit purposes so that may
react and remedy the outage. Credits are only due if the shortfall is
confirmed after 's investigation.
