peerShouldReceiveUpdate waited 500ms for the expected update message,
and every outer wrapper across the management/server test suite paired
it with a 1s goroutine-drain timeout. Both were too tight for slower
CI runners (MySQL, FreeBSD, loaded sqlite), producing intermittent
"Timed out waiting for update message" failures in tests like
TestDNSAccountPeersUpdate, TestPeerAccountPeersUpdate, and
TestNameServerAccountPeersUpdate.
Introduce peerUpdateTimeout (5s) next to the helper and use it both in
the helper and in every outer wrapper so the two timeouts stay in sync.
Only runs down on failure; passing tests return as soon as the channel
delivers, so there is no slowdown on green runs.
f732b01a05