[Feature Request] support multi-threaded encryption of trojan traffics
- Dominant language
- Go
- Stars
- 6.2k
- Forks
- 402
- Avg merge
- 57m
- Merged PRs (30d)
- 2
Description
### Greetings
_No response_
### Feature Request
With a `MT7986A` chip, with 750Mbps speed, `daed` (based on wing) runs out 80% of single core.
Switching to ShadowSocks nodes helps, but it requires more expensive proxy service for stablity.
One workaround can be: run another trojan implementation, must_direct it by pname,
or running trojan-go on some x86 computers
### Use Cases
Running dae on low-end devices, for example, routers
### Potential Benefits
Users could run dae(d) on low-end devices like JDCloud RE-CP-03, with 1Gbps trojan proxies, without an external proxy implementation
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by profiling dae's trojan traffic path on the reported MT7986A hardware and reading the existing wing-based implementation referenced in the issue. Compare the single-core bottleneck with the ShadowSocks and trojan-go alternatives. Done means trojan traffic can use multiple threads and achieves the intended throughput on low-end routers without requiring an external proxy implementation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- networking, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100