2024-11-25 00:25:03 +00:00
|
|
|
# cerulea-relay
|
|
|
|
|
|
2024-11-27 05:11:44 +00:00
|
|
|
Realtime non-archival relay for third-party AT Proto PDSes.
|
2024-11-25 00:25:03 +00:00
|
|
|
|
2024-11-27 05:11:44 +00:00
|
|
|
In the interest of cost control, we are scaling down the network:
|
|
|
|
|
- Only PDSes with fewer than 1000 repos are crawled
|
|
|
|
|
- We do no backfilling, only current events are relayed to consumers
|
|
|
|
|
- Stale data (≈ 24hrs?) is purged from the database [not doing this yet]
|
|
|
|
|
|
|
|
|
|
The idea is that we can have apps with much larger limits if we scale down the volume of the network.
|
2024-11-25 00:25:03 +00:00
|
|
|
- Large block sizes
|
|
|
|
|
- Large record size limit
|
2024-11-27 05:11:44 +00:00
|
|
|
- therefore: Large text field in post records, large uploads
|
2024-11-25 00:25:03 +00:00
|
|
|
- etcetcetc
|
2024-11-25 04:00:50 +00:00
|
|
|
|
|
|
|
|
## todo
|
|
|
|
|
|
|
|
|
|
- metrics / tracing / observability shit
|
|
|
|
|
- history:
|
2024-11-27 05:11:44 +00:00
|
|
|
- store indexedAt values
|
|
|
|
|
- purge based on ttl
|
|
|
|
|
- takedowns
|
