This patch introduces an adaptive dial queue that spawns a dynamically sized
set of goroutines to preemptively stage dials for later handoff to the DHT
protocol for RPC. It identifies backpressure on both ends (dial consumers and
dial producers), and takes compensating action by adjusting the worker pool.
We start with `DialQueueMinParallelism` number of workers (6), and scale up
and down based on demand and supply of dialled peers.
The following events trigger scaling:
- we scale up when we can't immediately return a successful dial to a new
consumer.
- we scale down when we've been idle for a while waiting for new dial
attempts.
- we scale down when we complete a dial and realise nobody was waiting for it.
Dialler throttling (e.g. FD limit exceeded) is a concern, as we can easily
spin up more workers to compensate, and end up adding fuel to the fire. Since
we have no deterministic way to detect this for now, we hard-limit concurrency
to `DialQueueMaxParallelism` (20).
Both, dht.go and dht_bootstrap.go duplicate the package documentation.
The whole block had been moved to doc.go
Also query.go provides documentation for query package but now it
belongs to dht package, so it's been removed.
Without this, one of the workers may hang when trying to re-grab the rate-limit
after dialing. I believe this may be what's causing some of our DHT requests to
stall (evidence: massive goroutine buildup on the gateways).
First, we use Alpha instead of K as the number of peers we grab from the
routing table (as per the kademlia paper).
Second, we don't use a size limited set for the 'GetClosestPeers' query.
We're going to process more than K peers before we find the K closest
peers.
Third, Change GetClosestPeers to actually return the K Closest peers,
not a hodge podge of peers that it found on the way to finding the
closest peers.
* Update golog in go-ipfs
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update go-libp2p for go-log
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update go-libp2p-secio for go-log
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update go-libp2p-crypto for go-log
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update go-libp2p-peer for go-log
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Import peersore, it wasn't imported
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update peerstore
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update peer
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update secio
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
* Update go-libp2p
License: MIT
Signed-off-by: Jakub Sztandera <kubuxu@protonmail.ch>
For the rest of the packages in util, move them to thirdparty
and update the references. util is gone!
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
Queries previously would sometimes only query three (alpha value) peers
before halting the operation. This PR changes the number of peers
grabbed from the routing table to start a query to K.
Dht nodes would also not respond with enough peers, as per the kademlia
paper, this has been changed to from 4 to 'K'.
The query mechanism itself also was flawed in that it would pull all the
peers it had yet to query out of the queue and 'start' the query for
them. The concurrency rate limiting was done inside the 'queryPeer'
method after the goroutine was spawned. This did not allow for peers
receiver from query replies to be properly queried in order of distance.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
