Bluesky, a rising decentralized social network, recently faced a major outage that lasted about an hour. Users on both web and mobile apps were unable to access the service during this time. The issue started around 6.55 PM ET and was linked to networking problems with its Personal Data Servers (PDS). A fix was started around 7.38 PM ET, and the platform was back shortly after.
This raised a common question – how can a decentralized platform go down?
The answer lies in how Bluesky is currently set up. While it’s built on the AT Protocol and is designed for decentralization, most users today rely on the main app and infrastructure run by Bluesky itself. In theory, anyone can host parts of the system like PDS and relays, but not many do right now because the platform is still in early stages.
Some users who had set up their own infrastructure were not affected by the outage. This shows the potential of true decentralization where each community or user can run their own version of the network. In the future, the plan is to have many apps and services built on the AT Protocol, each possibly managed by different teams.
For example, the Blacksky team is working on creating safer and more inclusive online spaces using decentralized tools. As more communities adopt these tools, the impact of any single failure like this outage would be much lower.
Until that future becomes a reality, Bluesky’s system outages can still impact most users.
The downtime also led to some playful rivalry. Mastodon, another decentralized social platform based on a different protocol called ActivityPub, saw users teasing Bluesky. One user humorously claimed their personal Raspberry Pi running Mastodon didn’t skip a beat during the Bluesky crash.
Even with the jokes and downtime, Bluesky was quick to fix the problem and restore access.
By
