cmd/utils: use eth DNS tree for snap discovery#22808
Conversation
When multiple iterators use the same enrtree:// URL on a shared client, duplicate DNS requests can be avoided using singleflight.Group.
Measurements have shown that > 75% of eth nodes in the public DNS trees support snap.
| eth.ethDialCandidates, err = setupDiscovery(eth.config.EthDiscoveryURLs) | ||
| // Setup DNS discovery iterators. | ||
| dnsclient := dnsdisc.NewClient(dnsdisc.Config{}) | ||
| eth.ethDialCandidates, err = dnsclient.NewIterator(eth.config.EthDiscoveryURLs...) |
There was a problem hiding this comment.
Out of curiosity, is this safe if eth.config.EthDiscoveryURLs == []? Previously we had a special clause that returned nil, nil for that, but here we're instantiating a random iterator even for no-urls. Just want a confirm that we won't spin loop shomewhere.
There was a problem hiding this comment.
But let me try it anyway.
There was a problem hiding this comment.
I've tested it, and there is no problem with empty list. The iterator is basically closed immediately.
| c.entries.Add(cacheKey, e) | ||
| return e, nil | ||
| }) | ||
| e, _ := ei.(entry) |
There was a problem hiding this comment.
Won't this crash if doResolveEntry errors? We're returning a nil interface in that case (line 184)
There was a problem hiding this comment.
This is why it uses the two-value type assertion. It would crash with e := ei.(entry).
This removes auto-configuration of the snap.*.ethdisco.net DNS discovery tree. Since measurements have shown that > 75% of nodes in all.*.ethdisco.net support snap, we have decided to retire the dedicated index for snap and just use the eth tree instead. The dial iterators of eth and snap now use the same DNS tree in the default configuration, so both iterators should use the same DNS discovery client instance. This ensures that the record cache and rate limit are shared. Records will not be requested multiple times. While testing the change, I noticed that duplicate DNS requests do happen even when the client instance is shared. This is because the two iterators request the tree root, link tree root, and first levels of the tree in lockstep. To avoid this problem, the change also adds a singleflight.Group instance in the client. When one iterator attempts to resolve an entry which is already being resolved, the singleflight object waits for the existing resolve call to finish and returns the entry to both places.
This removes auto-configuration of the
snap.*.ethdisco.netDNS discovery tree.Since measurements have shown that > 75% of nodes in
all.*.ethdisco.netsupportsnap, we have decided to retire the dedicated index for snap and just use the eth
tree instead.
The dial iterators of eth and snap now use the same DNS tree in the default configuration,
so both iterators should use the same DNS discovery client instance. This ensures that
the record cache and rate limit are shared. Records will not be requested multiple times.
While testing the change, I noticed that duplicate DNS requests do happen even
when the client instance is shared. This is because the two iterators request the tree
root, link tree root, and first levels of the tree in lockstep. To avoid this problem, the
change also adds a singleflight.Group instance in the client. When one iterator
attempts to resolve an entry which is already being resolved, the singleflight object
waits for the existing resolve call to finish and returns the entry to both places.