[Do not merge] Changes around channel management #2775
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…_channel_outgoing
|
eda_channels_1@29086 aka 20210714.9 vs main ewma over 20 builds from 28649 to 29064 Click to see table
|
|
2 more thoughts:
|
This was referenced Jul 8, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Opening this mega-PR to get CI testing, as the previously problematic tests are now passing locally. This includes many non-critical refactors/reworkings of adjacent bits of code, so I'll factor those into separate PRs.
A brief summary of the changes, as I remember them:
NodeToNodeChannelManagertakes some ofnode_state.h, and some ofchannels.h, and is the sole point for external systems to send/receive (they don't open/close/manage individual channels)associatemessage to the host and from then on the host can remember theID -> IPmapping, or else it is a response to an inbound message, in which case the host still has a single associate session. Then the host can create TCP sessions on-demand, re-creating them if they were closed by networking issues, etc, and just becomes another step in the best-effort-but-faulty network stack (WIP)node_outboundmessages over the ringbuffer. We previously benefitted from the fact that writing multiple values to the ringbuffer would blit them in a known format, which happened to match how we later read them withserialized::read(partly on the host, partly inside the receiving enclave). We now gives this explicit types like any other message, distinguishing the host-readable bits from the payload. There's a reserialisation penalty here but I think we can minimise it, and there's an associated fix to useByteRanges on these types that may mitigate its effect in practiceAnd the core goal: