This policy is part of the Terms of Service and works together with the Service Level Agreement. It explains what we replace, how quickly, and what you need to do.
1. Cloud servers (VPS)
If a VPS is unusable because of a fault on our side — a failed node, corrupted template, broken network configuration — we replace it at no charge:
- Where the disk is intact, the VM is migrated or restarted on another node in the same region, keeping its IP, data and billing.
- Where the disk is lost (node storage failure), we provision a fresh server of the same specification with the same IP and, if you have Joy backup space, restore the latest backup. Snapshots on the failed storage are lost, which is why they are not backups.
- Target time: within 4 hours of diagnosis for migrations; within 1 hour for a fresh server once loss is confirmed.
Faults inside the operating system — misconfiguration, malware, filled disks, broken updates — are not covered, but a reinstall is always free and support will help you diagnose.
2. Dedicated servers
Failed components (drives, memory, power supplies, NICs, mainboards) are replaced within 4 hours of diagnosis, 24/7, with equal or better parts. If a like-for-like part is unavailable we offer a temporary server of equal or greater specification while the original is repaired, or a permanent upgrade at the same price. Data on failed drives is not recoverable by Joy; RAID options and backup space are available for a reason. Where hardware fails repeatedly (three incidents in 90 days) you may request a different machine.
3. IP addresses
If an IP address assigned to you is unusable because of a problem attributable to Joy — routing errors, a listing on a major blocklist that predates your assignment, or a conflict — we replace it with a clean address at no charge and help you update rDNS and DNS. Listings caused by traffic from your server are your responsibility; we will guide you through delisting and may charge for repeated replacements.
4. Windows and other licences
If a licence add-on stops activating because of a problem on our side, we reissue it immediately. Licence problems caused by changes inside the OS are handled by a reinstall.
5. Hardware sold by Joy
New hardware sold by Joy carries a 12-month warranty against defects; refurbished hardware carries 90 days unless the listing says otherwise. Within the warranty period we repair or replace defective items at our cost, including shipping both ways in India. Outside India, return shipping is at our cost only where the item was dead on arrival. Damage from misuse, power events, environmental conditions or modification is excluded. To claim, open a Sales ticket with the invoice number, serial number and a description or photos of the fault; we issue a return authorisation within one business day.
6. Marketplace items
Replacement of items bought on the marketplace is between buyer and seller. For escrow transactions, a fault reported within 7 days of delivery pauses release of funds; the marketplace team decides between replacement, partial refund or full refund on the evidence provided.
7. Your responsibilities
- Report faults promptly by ticket with the server hostname or serial number, the symptoms and the time in UTC.
- Keep backups of your data; replacement restores capacity, not data, unless you use Joy backup space.
- Cooperate with diagnosis — for example by allowing a reboot or a console session — so that we can confirm the fault quickly.
8. Credits
Downtime caused by a fault covered here also counts towards SLA credits under the Service Level Agreement. Replacement and credit are complementary, not alternatives.