Why new mirrors were introduced
Over the past months, overall traffic to DRUGHUB has grown steadily, putting additional pressure on the original set of access nodes. While uptime remained high, latency spikes and occasional congestion windows showed that it was time to expand the mirror cluster and redistribute incoming connections more evenly.
The new mirrors are not “backup only” endpoints — they are fully integrated into the access layer. Load balancing and health checks ensure that users are routed to responsive nodes while failing or degraded instances are removed from rotation automatically.
What has changed technically
The updated infrastructure introduces multiple Tor v3 endpoints with hardened configs, isolated hosting zones and independent monitoring. Each mirror exposes the same DRUGHUB core, but runs behind separate network boundaries to limit the blast radius of any targeted attack.
Access policies remain the same: there are no clearnet frontends, no bridge domains and no mixed‑mode entry points. All connectivity stays strictly inside Tor, with onion URLs being the only supported channel.
How to use the new mirrors safely
Users should continue to treat onion URLs as sensitive credentials. Always obtain mirror addresses from PGP‑signed DRUGHUB announcements or from the official access hub inside the marketplace. Avoid trusting third‑party lists, screenshots or forwarded messages without cryptographic verification.
For operational security, rotate your Tor identity regularly, use dedicated wallets for marketplace activity and verify that the onion URL and PGP key fingerprint match the values published in official DRUGHUB materials.
What to expect going forward
The mirror set will be adjusted dynamically as load patterns evolve. New nodes may appear and old ones may be retired, but the underlying rule stays constant: any endpoint is considered valid only if it is referenced in a current PGP‑signed notice or on the internal access page.
From the user perspective the goal is simple: a familiar DRUGHUB interface, responsive at all hours, regardless of individual node status or external pressure on specific onion services.