Skip to content

[Bugfix] Do not Replace member if ScaleDown is required #939

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Mar 24, 2022
Merged

[Bugfix] Do not Replace member if ScaleDown is required #939

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@
- (Feature) (AT) Add ArangoTask API
- (Bugfix) Fix NPE in State fetcher
- (Refactor) Configurable throttle inspector
- (Bugfix) Skip Replace operation on DBServer if they need to be scaled down

## [1.2.8](https://github.com/arangodb/kube-arangodb/tree/1.2.8) (2022-02-24)
- Do not check License V2 on Community images
Expand Down
11 changes: 11 additions & 0 deletions pkg/deployment/reconcile/plan_builder_normal.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,6 +98,12 @@ func createMemberFailedRestorePlan(ctx context.Context,

// Check for members in failed state
status.Members.ForeachServerGroup(func(group api.ServerGroup, members api.MemberStatusList) error {
failed := 0
for _, m := range members {
if m.Phase == api.MemberPhaseFailed {
failed++
}
}
for _, m := range members {
if m.Phase != api.MemberPhaseFailed || len(plan) > 0 {
continue
Expand All @@ -112,6 +118,11 @@ func createMemberFailedRestorePlan(ctx context.Context,
continue
}

if c := spec.DBServers.GetCount(); c <= len(members)-failed {
// We have more or equal alive members than current count, we should not recreate this member
continue
}

if agencyState.Plan.Collections.IsDBServerInDatabases(m.ID) {
// DBServer still exists in agency plan! Will not be removed, but needs to be recreated
memberLog.Msg("Recreating DBServer - it cannot be removed gracefully")
Expand Down
50 changes: 48 additions & 2 deletions pkg/deployment/reconcile/plan_builder_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1058,13 +1058,13 @@ func TestCreatePlan(t *testing.T) {
ExpectedLog: "Creating member replacement plan because member has failed",
},
{
Name: "DBServer in failed state",
Name: "DBServer in failed state - recreate",
context: &testContext{
ArangoDeployment: deploymentTemplate.DeepCopy(),
},
Helper: func(ad *api.ArangoDeployment) {
ad.Spec.DBServers = api.ServerGroupSpec{
Count: util.NewInt(2),
Count: util.NewInt(3),
}
ad.Status.Members.DBServers[0].Phase = api.MemberPhaseFailed
ad.Status.Members.DBServers[0].ID = "id"
Expand All @@ -1076,6 +1076,26 @@ func TestCreatePlan(t *testing.T) {
},
ExpectedLog: "Creating member replacement plan because member has failed",
},
{
Name: "DBServer in failed state - remove",
context: &testContext{
ArangoDeployment: deploymentTemplate.DeepCopy(),
},
Helper: func(ad *api.ArangoDeployment) {
ad.Spec.DBServers = api.ServerGroupSpec{
Count: util.NewInt(2),
}
ad.Status.Members.DBServers[2].Phase = api.MemberPhaseFailed
ad.Status.Members.DBServers[2].ID = "id3"
},
ExpectedPlan: []api.Action{
actions.NewAction(api.ActionTypeCleanOutMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeKillMemberPod, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeShutdownMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeRemoveMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
},
ExpectedLog: "Creating scale-down plan",
},
{
Name: "CleanOut DBserver",
context: &testContext{
Expand All @@ -1101,6 +1121,32 @@ func TestCreatePlan(t *testing.T) {
},
ExpectedLog: "Creating dbserver replacement plan because server is cleanout in created phase",
},
{
Name: "CleanOut DBserver - scale down",
context: &testContext{
ArangoDeployment: deploymentTemplate.DeepCopy(),
},
Helper: func(ad *api.ArangoDeployment) {
ad.Spec.DBServers = api.ServerGroupSpec{
Count: util.NewInt(2),
}
ad.Status.Members.DBServers[2].ID = "id3"
ad.Status.Members.DBServers[2].Phase = api.MemberPhaseCreated
ad.Status.Members.DBServers[2].Conditions = api.ConditionList{
{
Type: api.ConditionTypeCleanedOut,
Status: core.ConditionTrue,
},
}
},
ExpectedPlan: []api.Action{
actions.NewAction(api.ActionTypeCleanOutMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeKillMemberPod, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeShutdownMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
actions.NewAction(api.ActionTypeRemoveMember, api.ServerGroupDBServers, withPredefinedMember("id3")),
},
ExpectedLog: "Creating scale-down plan",
},
{
Name: "Scale down DBservers",
context: &testContext{
Expand Down