Merge master to unstable (#570)

* Revisit Floodsub (#543)

Fixes #525

add coverage to unsubscribeAll and testing

* add mounted protos to identify message (#546)

* add stable/unstable auto bumps

* fix auto-bump CI

* merge nbc auto bump with CI in order to bump only on CI success

* put conditional locks on nbc bump (#549)

* Fix minor exception issues (#550)

Makes code compatible with
https://github.com/status-im/nim-chronos/pull/166 without requiring it.

* fix nimbus ref for auto-bump stable's PR

* use a builder pattern to build the switch (#551)

* use a builder pattern to build the switch

* with with

* more refs

* builders (#559)

* More builders (#560)

* address some issues pointed out in review

* re-add to prevent breaking other projects

* mem usage cleanups for pubsub (#564)

In `async` functions, a closure environment is created for variables
that cross an await boundary - this closure environment is kept in
memory for the lifetime of the associated future - this means that
although _some_ variables are no longer used, they still take up memory
for a long time.

In Nimbus, message validation is processed in batches meaning the future
of an incoming gossip message stays around for quite a while - this
leads to memory consumption peaks of 100-200 mb when there are many
attestations in the pipeline.

To avoid excessive memory usage, it's generally better to move non-async
code into proc's such that the variables therein can be released earlier
- this includes the many hidden variables introduced by macro and
template expansion (ie chronicles that does expensive exception
handling)

* move seen table salt to floodsub, use there as well
* shorten seen table salt to size of hash
* avoid unnecessary memory allocations and copies in a few places
* factor out message scoring
* avoid reencoding outgoing message for every peer
* keep checking validators until reject (in case there's both reject and
ignore)
* `readOnce` avoids `readExactly` overhead for single-byte read
* genericAssign -> assign2

* More gossip coverage (#553)

* add floodPublish test

* test delivery via control Iwant/have mechanics

* fix issues in control, and add testing

* fix possible backoff issue with pruned routine overriding it

* fix control messages (#566)

* remove unused control graft check in handleControl

* avoid sending empty Iwant messages

* Split dialer (#542)

* extracting dialing logic to dialer

* exposing upgrade methods on transport

* cleanup

* fixing tests to use new interfaces

* add comments

* add base exception class and fix hierarchy

* fix imports

* Merge master (#555)

* Revisit Floodsub (#543)

Fixes #525

add coverage to unsubscribeAll and testing

* add mounted protos to identify message (#546)

* add stable/unstable auto bumps

* fix auto-bump CI

* merge nbc auto bump with CI in order to bump only on CI success

* put conditional locks on nbc bump (#549)

* Fix minor exception issues (#550)

Makes code compatible with
https://github.com/status-im/nim-chronos/pull/166 without requiring it.

* fix nimbus ref for auto-bump stable's PR

* Split dialer (#542)

* extracting dialing logic to dialer

* exposing upgrade methods on transport

* cleanup

* fixing tests to use new interfaces

* add comments

* add base exception class and fix hierarchy

* fix imports

* `doAssert` is `ValueError` not `AssertionError`?

* revert back to `AssertionError`

Co-authored-by: Giovanni Petrantoni <7008900+sinkingsugar@users.noreply.github.com>
Co-authored-by: Jacek Sieka <jacek@status.im>

* Builders (#558)

* use a builder pattern to build the switch (#551)

* use a builder pattern to build the switch

* with with

* more refs

* Merge master (#555)

* Revisit Floodsub (#543)

Fixes #525

add coverage to unsubscribeAll and testing

* add mounted protos to identify message (#546)

* add stable/unstable auto bumps

* fix auto-bump CI

* merge nbc auto bump with CI in order to bump only on CI success

* put conditional locks on nbc bump (#549)

* Fix minor exception issues (#550)

Makes code compatible with
https://github.com/status-im/nim-chronos/pull/166 without requiring it.

* fix nimbus ref for auto-bump stable's PR

* Split dialer (#542)

* extracting dialing logic to dialer

* exposing upgrade methods on transport

* cleanup

* fixing tests to use new interfaces

* add comments

* add base exception class and fix hierarchy

* fix imports

* `doAssert` is `ValueError` not `AssertionError`?

* revert back to `AssertionError`

Co-authored-by: Giovanni Petrantoni <7008900+sinkingsugar@users.noreply.github.com>
Co-authored-by: Jacek Sieka <jacek@status.im>

* `doAssert` is `ValueError` not `AssertionError`?

* revert back to `AssertionError`

* fix builders

* more builder stuff

* more builders

Co-authored-by: Giovanni Petrantoni <7008900+sinkingsugar@users.noreply.github.com>
Co-authored-by: Jacek Sieka <jacek@status.im>

Co-authored-by: Giovanni Petrantoni <7008900+sinkingsugar@users.noreply.github.com>
Co-authored-by: Jacek Sieka <jacek@status.im>
This commit is contained in:
Dmitriy Ryajov 2021-05-06 14:41:49 -06:00
parent 5ee67f31bf
commit c949f14a99
No known key found for this signature in database
GPG Key ID: DA8C680CE7C657A4
1 changed files with 0 additions and 14 deletions

View File

@ -417,24 +417,10 @@ method publish*(g: GossipSub,
topic
trace "Publishing message on topic", data = data.shortLog
if topic.len <= 0: # data could be 0/empty
debug "Empty topic, skipping publish"
return 0
var peers: HashSet[PubSubPeer]
if g.parameters.floodPublish:
# With flood publishing enabled, the mesh is used when propagating messages from other peers,
# but a peer's own messages will always be published to all known peers in the topic.
for peer in g.gossipsub.getOrDefault(topic):
if peer.score >= g.parameters.publishThreshold:
trace "publish: including flood/high score peer", peer
peers.incl(peer)
# add always direct peers
peers.incl(g.explicit.getOrDefault(topic))
if topic in g.topics: # if we're subscribed use the mesh
peers.incl(g.mesh.getOrDefault(topic))
else: # not subscribed, send to fanout peers