17 November 2026 · Kryptonservers engineering
IPTV panel DDoS protection: the attacks panels get, and what stops each one
SYN floods on odd ports, player_api floods, credential stuffing, MAC cracking, account sharing and scanners: how each hits an IPTV panel and how to stop it.
An IPTV panel is an unusual thing to protect. It serves a public API that every player hits on a schedule, it listens on ports nobody else uses, and its users are spread across residential networks and shared addresses. Generic DDoS protection is built for websites; panels need more. These are the attacks panels get, and how each one is stopped.
1. SYN floods on the panel’s own port
Panels listen on 25461, 8080, 2082 or whatever the operator picked years ago. Many filters only protect 80 and 443, so a SYN flood on 25461 goes straight through.
What stops it: a SYN challenge on every port the panel listens on, applied automatically. In KryptonProxy the XDP program answers every new connection with a cookie, so forged sources never open a connection, whatever the port.
2. Request floods on player_api.php and get.php
Scripts hammer player_api.php or get.php with valid-looking requests. Each one makes the panel query its database. A few hundred requests a second from a handful of addresses can take a panel down without any network-level flood.
What stops it: per-address request counting at the proxy. Past the limit (600 requests a minute by default, tuned per panel) the address is banned, first in the proxy and then in the network card. In our tests the first flooding address was banned 0.24 seconds after the attack started.
3. Credential stuffing
One address tries hundreds of username and password pairs against the login endpoints: player_api.php, panel_api.php, get.php, xmltv.php.
What stops it: counting distinct usernames per address. A real player uses one account; an address trying more than three in a window is banned globally. In the lab, every one of 15 stuffing addresses was banned within 1.2 seconds.
4. MAC cracking on Stalker portals
For MAC-based portals, attackers walk through MAC address ranges looking for active devices.
What stops it: reading the MAC from portal requests (query or cookie) and counting distinct MACs per address. Walking a range shows up within a few requests.
5. One account, a hundred viewers
Account sharing is not an attack on the network, but it costs real money and load: one line, used from far more addresses than any household has.
What stops it: tracking the addresses each account is used from, and blocking the account for a while when it crosses the limit, instead of letting it degrade service for everyone.
6. Scanners
Bots looking for admin pages, backups and known exploits collect 404s and fill logs.
What stops it: scoring 404s and error rates per address. A scanner is banned within seconds of its first burst; your logs stay readable.
Putting it together
These checks have to run on every request without slowing anything down, and the bans have to land where they cost nothing: in the network card. In our replay of a mixed attack (50 attacking addresses across all six patterns, 100 normal players, and a trusted source at 196,000 requests a second) every attacker was banned, no player was, and the players’ slowest 1 % of requests stayed under 6 ms.
Running a panel that gets hit? Tell us about it or see the IPTV protection page.