akovalenko/lnd.xprv
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Revert "test: account for block race by mining additional block in remote htlc force close test"

Browse Source 
Before the previous commit, we assumed the HTLC's timeout transaction
would be the only transaction in the mempool. In reality, after mining
some blocks for the HTLC to expire and waiting for the timeout
transaction to arrive in the mempool, at times we would instead detect
the funding output's sweeping transaction and proceed the test with this
assumption, leading to the case where we would have to mine extra blocks
to include the HTLC sweeping transaction. This has been resolved in the
previous commit, so this fix is no longer needed.

This reverts commit e54f1ea4db.
This commit is contained in:
Wilmer Paulino 2018-05-24 22:55:13 -04:00
parent 84fcf5b12f
commit 3e7c69e596
No known key found for this signature in database
GPG Key ID: 6DF57B9F9514972F
1 changed files with 2 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
lnd_test.go
View File

@ -7127,20 +7127,9 @@ func testMultiHopRemoteForceCloseOnChainHtlcTimeout(net *lntest.NetworkHarness,
nodes = []*lntest.HarnessNode{net.Alice}
err = lntest.WaitPredicate(func() bool {
return assertNumActiveHtlcs(nodes, 0)
}, time.Second*8)
}, time.Second*15)
if err != nil {
// It may be the case that due to a race, Bob's sweeping
// transaction hasn't yet been confirmed, so we'll mine another
// block to nudge it in. If after this it still Alice will has
// an HTLC, then it's actually a test failure.
if _, err := net.Miner.Node.Generate(1); err != nil {
t.Fatalf("unable to generate block: %v", err)
}
if err = lntest.WaitPredicate(func() bool {
return assertNumActiveHtlcs(nodes, 0)
}, time.Second*8); err != nil {
t.Fatalf("alice's channel still has active htlc's")
}
t.Fatalf("alice's channel still has active htlc's")
}
// Now we'll check Bob's pending channel report. Since this was Carol's