CCF

CCF copied to clipboard

Stop keeping track of destroyed node-to-node channels

1

We currently keep track of destroyed node-to-node channels (i.e. replace the `std::shared_ptr` with `nullptr` in the channels map, instead of simply deleting the `NodeId` key). We should try to fix,...

jumaffre

Destroy idle node-to-node channels automatically

2

Our current policy to closing node-to-node channels was originally too strict: a channel to a node is destroyed as soon as we have (globally) committed its retirement. #2654 relaxes that,...

jumaffre

CCF node should reply with a timeout error if a request takes too long to execute

4

In the case of `Write` requests executed directly on the leader or `Read` on any node, RPC requests execute and return synchronously to the caller. However, in the case of...

jumaffre

Ability to reject requests if too many transactions pending

1

Whenever too many transactions are PENDING, then there is the risk of running out of memory since those transactions are kept in memory. Getting into this state may be because...

letmaik

JS Runtime execution efficiency and sandboxing improvements

6

Performance items: - [x] ~~Share runtime across execution contexts? Per-session?~~ (experiment: https://github.com/microsoft/CCF/pull/2162) - [x] Investigate performance benefit from caching pre-compiled bytecode? Implemented in #2643. Sandboxing/Fairness: - [ ] Expose context...

achamayou

We may want to look at logging libraries such as https://github.com/odygrd/quill, on platforms that aren't as restricted as SGX. _Originally posted by @achamayou in https://github.com/microsoft/CCF/issues/4013#issuecomment-1185863405_

eddyashton

Add CCF benchmark to TechEmpower/FrameworkBenchmarks

1

Now that we support HTTP (i.e. non-HTTPS) interfaces, we could add CCF to https://github.com/TechEmpower/FrameworkBenchmarks so that we have a reputable benchmark of our HTTP server implementation.

jumaffre

Using keys with a passphrase seems to lead to invalid signatures. To be reproduced and investigated.

achamayou

Prevent nodes added during a rolled back reconfiguration from disrupting the service

1

It is my understanding that newly added nodes begin ticking (see: https://github.com/microsoft/CCF/blob/src/consensus/aft/raft.h#L1305) and thus can become a candidate when their reconfiguration txn becomes committable (followed by a signature txn). If...

heidihoward

Re-work TLS to match OpenSSL's style

1

After #3361 and #3394, there's only an OpenSSL implementation of TLS in CCF. However, because we have MbedTLS as the previous implementation, and we wanted to play safe on the...

rengolin