40 lines
1.3 KiB
Markdown
40 lines
1.3 KiB
Markdown
|
# 0005. Use Discovery v5 protocol
|
||
|
|
|
||
|
|
| Date | Tags |
|
||
|
|
|---|---|
|
||
|
|
| 2018-05-14 | architecture, p2p, ethereum |
|
||
|
|
|
||
|
|
|
||
|
|
## Status
|
||
|
|
|
||
|
|
proposed
|
||
|
|
|
||
|
|
## Context
|
||
|
|
|
||
|
|
Currently peers for messaging are hardcoded in the app. This means we can't
|
||
|
|
easily scale, as some of these nodes might fill up when we have more than N
|
||
|
|
active users.
|
||
|
|
|
||
|
|
Discovery v5 protocol (https://ideas.status.im/ideas/092-disc-v5-research,
|
||
|
|
https://github.com/fjl/p2p-drafts/blob/master/discv5-eip.md) allow us to
|
||
|
|
discover new peers from the app and scale server peers in our cluster
|
||
|
|
independently. This is vital for high availability/reliability and security.
|
||
|
|
|
||
|
|
## Decision
|
||
|
|
|
||
|
|
- Use and enable Discovery v5 for all networks
|
||
|
|
- Do this before beta is released
|
||
|
|
|
||
|
|
This means we have capacity to achieve 5000+ DAU / 500 concurrent connection. It
|
||
|
|
seems unlikely we can do so with reliable messaging in the current setup.
|
||
|
|
|
||
|
|
Trade-off for this is:
|
||
|
|
- Uncertainty in introducing a new protocol leading to possible delays
|
||
|
|
- Possible increase in CPU/Network (we can disable protocol once peers are found though)
|
||
|
|
|
||
|
|
## Consequences
|
||
|
|
|
||
|
|
- This is a top priority for P2P team, lead by @dshulyak and assisted by @adambabik and @jacob
|
||
|
|
- We need this battle tested to get it into app as soon as possible
|
||
|
|
- Chat, mobile team and QA team should be aware of this protocol change
|