mirror of
https://github.com/openziti/ziti.git
synced 2026-10-27 05:57:06 +00:00
929ea54eaf
BuildRouterLinks runs once per router connect and scanned the whole link table to find the connecting router's links. The table is proportional to the square of the router count in a full mesh, so a wave in which every router reconnects walks the cube of it in link visits. That never shows in steady state; it arrives during a mass reconnect, when the connect path is already the bottleneck. Indexing every link under both endpoint router ids fixes the cost, and introduces the problem that the table and the index are two structures written in sequence. Three things follow from that, and the destination repair below turned out to depend on the connect path's ordering as well. Performance: - indexes every link under both endpoint router ids, maintained in the add and remove paths all link mutations funnel through, so a connect costs its own router's link count rather than the whole table - prunes a full refresh against the reporting router's own links. That report is what every reconnect sends, and it walked the whole table to find the router's stale links, so the cost the connect path shed was still being paid one message later - indexes link ids rather than link pointers, so a per-router query resolves each id against the table and hands back whatever it currently holds. A reconnect replaces a link's Router object and a higher iteration replaces the Link object, but neither replaces the id, which is exactly why the scan existed - holds each router's link ids in a LockedSet rather than a sharded map, which cost a map and a mutex per shard for an index that is usually empty and is held for as long as the router has existed: 2.9KB per router id against 150 bytes - drops a router's index when the router is deleted, guarded on the store still having no router under that id and checked inside the index map's removal callback. A router id can come back, since fabric router ids are the enrollment certificate's common name, and nothing orders the cleanup against that: store event handlers run after bolt has released the writer lock, so one can execute after the id was recreated, connected and reported links. Indexing a link takes the same shard as the guarded read, so an entry for a live router cannot appear without its create having committed first, which the read then sees, and an entry arriving later blocks and lands in a freshly created index - keeps an index when the store cannot be read, since that is not evidence the router is gone, and dropping one otherwise could discard a link another goroutine had already fetched the index to record Keeping the two structures agreeing: - serializes Add and Remove on the link's id. Without it a removal running between an add's two writes takes the table entry, finds nothing yet in the index, and lets the add index a link the table no longer holds, which nothing cleans up. Only RouterReportedLink held that lock before; the four other removal paths held nothing - serializes the connect-time pairing on that id as well, which was the last writer left outside it. A removal reading no destination could complete while the pairing was still deciding, leaving the link in the destination's link set after the table had dropped it, where path computation still routed over it - unindexes an id only when the table no longer holds a link under it, so a replacement that has taken the table slot is not unindexed by the older link's removal. A stale id is the safe direction to err in, since a query filters it, where an id wrongly dropped leaves a live link no per-router query finds - has per-router queries resolve ids against the table, so an id it no longer holds is not handed out as a live link Pairing a link with its destination: - re-resolves the destination once the link is in the table, and on every later report, so a link that reached a bad state heals. A report resolves the destination before the link is in the table and a router's connect repairs links already in it, so a router connecting between those two moments was seen by neither, leaving a link with no adjacency that every operator-facing view still called healthy - makes that re-resolve unconditional rather than only for a link holding no destination. The resolve happens under the source router's connect stripe, not the destination's, so the destination can be replaced between the resolve and the report landing, and a link left on a connection that is no longer registered carries no adjacency just as one holding none does - registers a connecting router before building its links, which is what makes that re-resolve close the window rather than narrow it - adds Link.PointDestAt so the two repair paths cannot both index the link on the destination, since that index is a slice that does not deduplicate Also here: - logs the link count on both sides of the reconnect exchange, and why a report was discarded. A router announces its links once per reconnect and is not asked again, so a discarded report left it with no links for a router that believed it had announced them, with nothing recorded to compare. The discard reports the state it decided on, read once, since a reconnect landing between the decision and the log would otherwise have it name a state that discards nothing - gives test routers version info, which a connected router always has because the accept path refuses a hello without it, and which the router sync path dereferences with no nil check The destination repair is backported from the gossip link-state branch, where the connect path registers before building links for unrelated reasons. That ordering is load-bearing here and is now explicit at the call site.
470 lines
16 KiB
Go
470 lines
16 KiB
Go
/*
|
|
Copyright NetFoundry Inc.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
https://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
*/
|
|
|
|
package model
|
|
|
|
import (
|
|
"fmt"
|
|
"sync"
|
|
"sync/atomic"
|
|
"testing"
|
|
"time"
|
|
|
|
"github.com/openziti/ziti/v2/common/pb/ctrl_pb"
|
|
"github.com/stretchr/testify/assert"
|
|
"github.com/stretchr/testify/require"
|
|
)
|
|
|
|
// A simple test to check for failure of alignment on atomic operations for 64 bit variables in a struct
|
|
func Test64BitAlignment(t *testing.T) {
|
|
defer func() {
|
|
if r := recover(); r != nil {
|
|
t.Errorf("One of the variables that was tested is not properly 64-bit aligned.")
|
|
}
|
|
}()
|
|
|
|
link := Link{}
|
|
|
|
atomic.LoadInt64(&link.SrcLatency)
|
|
atomic.LoadInt64(&link.DstLatency)
|
|
atomic.LoadInt64(&link.Cost)
|
|
}
|
|
|
|
func TestLifecycle(t *testing.T) {
|
|
linkController := NewLinkManager(nil)
|
|
|
|
r0 := NewRouter("r0", "", "", 0, true)
|
|
r1 := NewRouter("r1", "", "", 0, true)
|
|
l0 := &Link{
|
|
Id: "l0",
|
|
Src: r0,
|
|
DstId: r1.Id,
|
|
}
|
|
l0.Dst.Store(r1)
|
|
|
|
linkController.Add(l0)
|
|
assert.True(t, linkController.has(l0))
|
|
|
|
links := r0.routerLinks.GetLinks()
|
|
assert.Equal(t, 1, len(links))
|
|
assert.Equal(t, l0, links[0])
|
|
|
|
links = r1.routerLinks.GetLinks()
|
|
assert.Equal(t, 1, len(links))
|
|
assert.Equal(t, l0, links[0])
|
|
|
|
linkController.Remove(l0)
|
|
assert.False(t, linkController.has(l0))
|
|
|
|
links = r0.routerLinks.GetLinks()
|
|
assert.Equal(t, 0, len(links))
|
|
|
|
links = r1.routerLinks.GetLinks()
|
|
assert.Equal(t, 0, len(links))
|
|
}
|
|
|
|
func TestNeighbors(t *testing.T) {
|
|
linkController := NewLinkManager(nil)
|
|
|
|
r0 := NewRouterForTest("r0", "", nil, nil, 0, true)
|
|
r1 := NewRouterForTest("r1", "", nil, nil, 0, true)
|
|
l0 := NewTestLink("l0", r0, r1)
|
|
l0.SetState(Connected)
|
|
linkController.Add(l0)
|
|
|
|
neighbors := linkController.ConnectedNeighborsOfRouter(r0)
|
|
assert.Equal(t, 1, len(neighbors))
|
|
assert.Equal(t, r1, neighbors[0])
|
|
}
|
|
|
|
// Test_RouterReportedLink_serializesAndStableOnStaleReport covers two
|
|
// RouterReportedLink guarantees: concurrent reports for the same link serialize
|
|
// through the striped per-link lock (no corruption under -race, and the highest
|
|
// reported iteration wins regardless of arrival order), and a stale/same-iteration
|
|
// report is a no-op that returns the existing link without replacing it or its
|
|
// source/dest routers.
|
|
func Test_RouterReportedLink_serializesAndStableOnStaleReport(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
src.Connected.Store(true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
dst.Connected.Store(true)
|
|
|
|
report := func(iteration uint32) *ctrl_pb.RouterLinks_RouterLink {
|
|
return &ctrl_pb.RouterLinks_RouterLink{
|
|
Id: "l0",
|
|
DestRouterId: dst.Id,
|
|
LinkProtocol: "tls",
|
|
DialAddress: "tcp:localhost:1234",
|
|
Iteration: iteration,
|
|
}
|
|
}
|
|
|
|
link, created := lm.RouterReportedLink(report(1), src, dst)
|
|
req.True(created)
|
|
req.NotNil(link)
|
|
req.Same(src, link.Src)
|
|
req.Same(dst, link.GetDest())
|
|
|
|
// A stale/same-iteration report returns the existing link unchanged: created is
|
|
// false, the same link object comes back, and the source router is not swapped
|
|
// even though a different (also connected) router object with the same id reports.
|
|
otherSrc := NewRouter("r0", "", "", 0, true)
|
|
otherSrc.Connected.Store(true)
|
|
again, created2 := lm.RouterReportedLink(report(1), otherSrc, dst)
|
|
req.False(created2)
|
|
req.Same(link, again, "same-iteration report returns the existing link")
|
|
req.Same(src, again.Src, "same-iteration report must not replace the source router")
|
|
|
|
// Concurrent reports for the same link serialize through the per-link lock. With
|
|
// mixed iterations arriving in arbitrary order the highest iteration must win
|
|
// (lower ones are rejected as stale once it lands), and the table holds exactly
|
|
// one link. Run under -race to catch unsynchronized access.
|
|
const maxIteration = 32
|
|
var wg sync.WaitGroup
|
|
for i := uint32(1); i <= maxIteration; i++ {
|
|
i := i
|
|
wg.Add(1)
|
|
go func() {
|
|
defer wg.Done()
|
|
lm.RouterReportedLink(report(i), src, dst)
|
|
}()
|
|
}
|
|
wg.Wait()
|
|
|
|
got, ok := lm.Get("l0")
|
|
req.True(ok)
|
|
req.Equal(uint32(maxIteration), got.Iteration, "highest reported iteration wins regardless of arrival order")
|
|
req.Len(lm.All(), 1, "exactly one link remains in the table")
|
|
}
|
|
|
|
// Test_linkTable_forRouter asserts a link is indexed under both of its endpoints, and unindexed from both
|
|
// when removed.
|
|
func Test_linkTable_forRouter(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
r0 := NewRouterForTest("r0", "", nil, nil, 0, true)
|
|
r1 := NewRouterForTest("r1", "", nil, nil, 0, true)
|
|
r2 := NewRouterForTest("r2", "", nil, nil, 0, true)
|
|
|
|
l01 := NewTestLink("l01", r0, r1)
|
|
l02 := NewTestLink("l02", r0, r2)
|
|
lm.Add(l01)
|
|
lm.Add(l02)
|
|
|
|
req.ElementsMatch([]*Link{l01, l02}, lm.linkTable.forRouter("r0"), "indexed under the source router")
|
|
req.Equal([]*Link{l01}, lm.linkTable.forRouter("r1"), "indexed under the destination router")
|
|
req.Equal([]*Link{l02}, lm.linkTable.forRouter("r2"))
|
|
req.Empty(lm.linkTable.forRouter("no-such-router"))
|
|
|
|
lm.Remove(l01)
|
|
req.Equal([]*Link{l02}, lm.linkTable.forRouter("r0"), "removal unindexes from the source")
|
|
req.Empty(lm.linkTable.forRouter("r1"), "removal unindexes from the destination")
|
|
}
|
|
|
|
// Test_linkTable_indexStaysConsistentUnderChurn is the leak guard for the router-id index: after a mesh is
|
|
// built, partly torn down, and has links replaced by higher iterations, the index must hold exactly what the
|
|
// table holds for every router. A change that altered a link's endpoint ids after it was added, or that
|
|
// bypassed the add/remove path, would strand entries here and keep dead links reachable.
|
|
func Test_linkTable_indexStaysConsistentUnderChurn(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
const routerCount = 12
|
|
routers := make([]*Router, routerCount)
|
|
for i := range routers {
|
|
routers[i] = NewRouterForTest(fmt.Sprintf("r%02d", i), "", nil, nil, 0, true)
|
|
}
|
|
|
|
linkId := func(i, j int) string { return fmt.Sprintf("l-%02d-%02d", i, j) }
|
|
|
|
// A full mesh, so every router carries links at both ends.
|
|
for i := 0; i < routerCount; i++ {
|
|
for j := i + 1; j < routerCount; j++ {
|
|
lm.Add(NewTestLink(linkId(i, j), routers[i], routers[j]))
|
|
}
|
|
}
|
|
|
|
// Remove a scattering of links.
|
|
for i := 0; i < routerCount; i += 3 {
|
|
for j := i + 1; j < routerCount; j += 2 {
|
|
if link, found := lm.Get(linkId(i, j)); found {
|
|
lm.Remove(link)
|
|
}
|
|
}
|
|
}
|
|
|
|
// Replace some links with a higher iteration, which removes and re-adds the same link id.
|
|
for i := 1; i < routerCount; i += 4 {
|
|
j := i + 1
|
|
lm.RouterReportedLink(&ctrl_pb.RouterLinks_RouterLink{
|
|
Id: linkId(i, j),
|
|
DestRouterId: routers[j].Id,
|
|
LinkProtocol: "tls",
|
|
DialAddress: "tcp:localhost:1234",
|
|
Iteration: 7,
|
|
}, routers[i], routers[j])
|
|
}
|
|
|
|
// Connect-time repointing replaces the Router objects a link references. The index is keyed on ids, so
|
|
// this must not disturb it.
|
|
for i := range routers {
|
|
fresh := NewRouterForTest(routers[i].Id, "", nil, nil, 0, true)
|
|
lm.BuildRouterLinks(fresh)
|
|
}
|
|
|
|
// The index must agree with the table, per router, in both directions.
|
|
for _, router := range routers {
|
|
expected := map[string]*Link{}
|
|
for _, link := range lm.All() {
|
|
if link.Src.Id == router.Id || link.DstId == router.Id {
|
|
expected[link.Id] = link
|
|
}
|
|
}
|
|
|
|
indexed := map[string]*Link{}
|
|
for _, link := range lm.linkTable.forRouter(router.Id) {
|
|
indexed[link.Id] = link
|
|
}
|
|
|
|
req.Equal(expected, indexed, "index and table disagree for router %v", router.Id)
|
|
}
|
|
}
|
|
|
|
// Test_linkTable_removeKeepsAReplacementIndexed covers the guard on index cleanup: a removal must not
|
|
// unindex an id the table holds a link under again. Add and Remove serialize on the link id, so this is
|
|
// belt-and-braces, kept because dropping the id is invisible in the table and surfaces only when a
|
|
// per-router query later fails to find the link.
|
|
func Test_linkTable_removeKeepsAReplacementIndexed(t *testing.T) {
|
|
req := require.New(t)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
|
|
lt := newLinkTable()
|
|
link := NewTestLink("l0", src, dst)
|
|
lt.add(link)
|
|
|
|
// The interleave the guard exists for: a replacement takes the table slot and indexes itself between a
|
|
// removal's table write and its unindex. Driven through unindex directly, since remove's iteration
|
|
// check refuses the primary removal once the replacement is in the table.
|
|
replacement := NewTestLink("l0", src, dst)
|
|
replacement.Iteration = link.Iteration + 1
|
|
lt.links.Set(replacement.Id, replacement)
|
|
|
|
lt.unindex(link)
|
|
|
|
for _, routerId := range []string{src.Id, dst.Id} {
|
|
req.Equal([]*Link{replacement}, lt.forRouter(routerId),
|
|
"the replacement must stay reachable under %v", routerId)
|
|
}
|
|
|
|
// With nothing live under the id, the same call must drop it rather than strand it. forRouter cannot
|
|
// tell the two apart, so this asserts against the index.
|
|
lt.links.Remove(replacement.Id)
|
|
lt.unindex(link)
|
|
|
|
for _, routerId := range []string{src.Id, dst.Id} {
|
|
index, found := lt.byRouter.Get(routerId)
|
|
req.True(found)
|
|
req.False(index.Contains(link.Id), "a removed link's id must not be left indexed under %v", routerId)
|
|
}
|
|
}
|
|
|
|
// Test_linkTable_forRouterTreatsTheTableAsAuthoritative asserts a per-router query does not hand back a link
|
|
// the table no longer holds. The index is written in a separate step, so it can briefly disagree.
|
|
func Test_linkTable_forRouterTreatsTheTableAsAuthoritative(t *testing.T) {
|
|
req := require.New(t)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
|
|
lt := newLinkTable()
|
|
link := NewTestLink("l0", src, dst)
|
|
lt.add(link)
|
|
req.Len(lt.forRouter(src.Id), 1)
|
|
|
|
// A stale index entry: gone from the table, still indexed.
|
|
lt.links.Remove(link.Id)
|
|
|
|
req.Empty(lt.forRouter(src.Id), "a link the table no longer holds must not be returned")
|
|
req.Empty(lt.forRouter(dst.Id), "nor under its other endpoint")
|
|
|
|
// The same id resolving to a new link is not staleness, and must be returned.
|
|
replacement := NewTestLink("l0", src, dst)
|
|
replacement.Iteration = link.Iteration + 1
|
|
lt.links.Set(replacement.Id, replacement)
|
|
|
|
req.Equal([]*Link{replacement}, lt.forRouter(src.Id), "a replaced link resolves to whatever the table holds")
|
|
req.Equal([]*Link{replacement}, lt.forRouter(dst.Id))
|
|
}
|
|
|
|
// Test_LinkManager_BuildRouterLinksSerializesOnLinkId asserts the connect-time pairing takes the same
|
|
// stripe every other change to a link takes. Without it a removal that reads no destination can complete
|
|
// between this reading the table and adding to the router's link set, leaving the link in that set after
|
|
// the table has dropped it, where path computation still routes over it.
|
|
func Test_LinkManager_BuildRouterLinksSerializesOnLinkId(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
link := NewTestLink("l0", src, dst)
|
|
link.Dst.Store(nil)
|
|
lm.Add(link)
|
|
|
|
// Stands in for a removal of this link already in progress.
|
|
unlock := lm.linkLocks.LockFor(link.Id)
|
|
|
|
built := make(chan struct{})
|
|
go func() {
|
|
lm.BuildRouterLinks(dst)
|
|
close(built)
|
|
}()
|
|
|
|
select {
|
|
case <-built:
|
|
req.Fail("the link build paired a link while another change to it was in flight")
|
|
case <-time.After(50 * time.Millisecond):
|
|
}
|
|
|
|
unlock()
|
|
|
|
select {
|
|
case <-built:
|
|
case <-time.After(5 * time.Second):
|
|
req.Fail("the link build did not proceed once the link's stripe was released")
|
|
}
|
|
req.Same(dst, link.GetDest(), "the link must be paired once the stripe is free")
|
|
req.Len(dst.GetLinks(), 1)
|
|
}
|
|
|
|
// Test_LinkManager_BuildRouterLinksSkipsUnresolvedIds: the index can name an id the table no longer holds,
|
|
// which is the state a removal leaves behind when it keeps an id a replacement may have taken. The pairing
|
|
// resolves ids under the stripe and must skip the ones that resolve to nothing, rather than adding a link
|
|
// the table has dropped to the router's link set, where nothing would clean it up.
|
|
func Test_LinkManager_BuildRouterLinksSkipsUnresolvedIds(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
link := NewTestLink("l0", src, dst)
|
|
link.Dst.Store(nil)
|
|
lm.Add(link)
|
|
|
|
// Gone from the table, still indexed.
|
|
lm.linkTable.links.Remove(link.Id)
|
|
req.Equal([]string{link.Id}, lm.linkTable.linkIdsForRouter(dst.Id), "the id must still be indexed")
|
|
|
|
lm.BuildRouterLinks(dst)
|
|
|
|
req.Nil(link.GetDest(), "a link the table no longer holds must not be paired")
|
|
req.Empty(dst.GetLinks(), "nor added to the router's link set")
|
|
}
|
|
|
|
// Test_LinkManager_RouterDeletedDropsTheIndex: a router id is only ever added to the index, so without a
|
|
// drop on delete every router that has ever had a link is held for the controller's lifetime. The links
|
|
// themselves are left alone, since a deleted router's peers still hold theirs.
|
|
func Test_LinkManager_RouterDeletedDropsTheIndex(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
gone := NewRouter("r0", "", "", 0, true)
|
|
peer := NewRouter("r1", "", "", 0, true)
|
|
link := NewTestLink("l0", gone, peer)
|
|
lm.Add(link)
|
|
|
|
req.Len(lm.LinksForRouter(gone.Id), 1)
|
|
req.Len(lm.LinksForRouter(peer.Id), 1)
|
|
|
|
lm.RouterDeleted(gone.Id)
|
|
|
|
_, found := lm.linkTable.byRouter.Get(gone.Id)
|
|
req.False(found, "the deleted router's index must be dropped, not just emptied")
|
|
req.Empty(lm.LinksForRouter(gone.Id))
|
|
|
|
req.Len(lm.LinksForRouter(peer.Id), 1, "the other endpoint's index must be untouched")
|
|
req.True(lm.has(link), "dropping an index must not remove the link itself")
|
|
|
|
// A report still in flight recreates the entry. Nothing reads it, since a deleted router cannot connect.
|
|
lm.Add(link)
|
|
req.Len(lm.LinksForRouter(peer.Id), 1, "the peer's index must not be double-counted by the recreate")
|
|
}
|
|
|
|
// Test_LinkManager_AddAndRemoveSerializeOnLinkId asserts a change to a link cannot begin while another
|
|
// change to the same link is in flight. That is what keeps a removal from running between an add's table and
|
|
// index writes, which would leave the link indexed but untabled. The window is two map writes wide, so the
|
|
// property is asserted directly rather than by stressing it.
|
|
func Test_LinkManager_AddAndRemoveSerializeOnLinkId(t *testing.T) {
|
|
req := require.New(t)
|
|
lm := NewLinkManager(nil)
|
|
|
|
src := NewRouter("r0", "", "", 0, true)
|
|
dst := NewRouter("r1", "", "", 0, true)
|
|
link := NewTestLink("l0", src, dst)
|
|
|
|
// Stands in for a change to this link already in progress.
|
|
unlock := lm.linkLocks.LockFor(link.Id)
|
|
|
|
added := make(chan struct{})
|
|
go func() {
|
|
lm.Add(link)
|
|
close(added)
|
|
}()
|
|
|
|
select {
|
|
case <-added:
|
|
req.Fail("Add published while another change to the same link was in flight")
|
|
case <-time.After(50 * time.Millisecond):
|
|
}
|
|
|
|
unlock()
|
|
|
|
select {
|
|
case <-added:
|
|
case <-time.After(5 * time.Second):
|
|
req.Fail("Add did not proceed once the link's stripe was released")
|
|
}
|
|
req.True(lm.has(link))
|
|
|
|
// Remove takes the same stripe, so it is serialized against an add for the same link either way round.
|
|
unlock = lm.linkLocks.LockFor(link.Id)
|
|
removed := make(chan struct{})
|
|
go func() {
|
|
lm.Remove(link)
|
|
close(removed)
|
|
}()
|
|
|
|
select {
|
|
case <-removed:
|
|
req.Fail("Remove retracted while another change to the same link was in flight")
|
|
case <-time.After(50 * time.Millisecond):
|
|
}
|
|
|
|
unlock()
|
|
|
|
select {
|
|
case <-removed:
|
|
case <-time.After(5 * time.Second):
|
|
req.Fail("Remove did not proceed once the link's stripe was released")
|
|
}
|
|
req.False(lm.has(link))
|
|
}
|