Skip to content

[7.0.3] healthcheck: use isIncluded correctly to fix replica/rdonly routing bug #6905

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
askdba merged 2 commits into from
Oct 22, 2020
Merged

[7.0.3] healthcheck: use isIncluded correctly to fix replica/rdonly routing bug #6905

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
78ac16f
healthcheck: use isIncluded correctly to fix replica/rdonly routing bug
deepthi Oct 20, 2020
a230073
fix comments
deepthi Oct 21, 2020
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
27 changes: 15 additions & 12 deletions go/vt/discovery/healthcheck.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -386,8 +386,8 @@ func (hc *HealthCheckImpl) deleteTablet(tablet *topodata.Tablet) {
log.Infof("We have no health data for tablet: %v, it might have been deleted already", tabletAlias)
return
}
// calling this will end the context associated with th.checkConn
// which will call finalizeConn, which will close the connection
// Calling this will end the context associated with th.checkConn,
// which will call finalizeConn, which will close the connection.
th.cancelFunc()
delete(hc.healthByAlias, tabletAlias)
// delete from map by keyspace.shard.tabletType
Expand Down Expand Up @@ -442,24 +442,27 @@ func (hc *HealthCheckImpl) updateHealth(th *TabletHealth, shr *query.StreamHealt
}
}
if !trivialNonMasterUpdate {
if shr.Target.TabletType != topodata.TabletType_MASTER {
// We re-sort the healthy tablet list whenever we get a health update for tablets we can route to.
// Tablets from other cells for non-master targets should not trigger a re-sort;
// they should also be excluded from healthy list.
if shr.Target.TabletType != topodata.TabletType_MASTER && hc.isIncluded(shr.Target.TabletType, shr.TabletAlias) {
all := hc.healthData[targetKey]
allArray := make([]*TabletHealth, 0, len(all))
for _, s := range all {
// only tablets in same cell / cellAlias are included in healthy list
if hc.isIncluded(shr) {
// Only tablets in same cell / cellAlias are included in healthy list.
if hc.isIncluded(s.Tablet.Type, s.Tablet.Alias) {
allArray = append(allArray, s)
}
}
hc.healthy[targetKey] = FilterStatsByReplicationLag(allArray)
}
if targetChanged && currentTarget.TabletType != topodata.TabletType_MASTER { // also recompute old target's healthy list
if targetChanged && currentTarget.TabletType != topodata.TabletType_MASTER && hc.isIncluded(shr.Target.TabletType, shr.TabletAlias) { // also recompute old target's healthy list
oldTargetKey := hc.keyFromTarget(currentTarget)
all := hc.healthData[oldTargetKey]
allArray := make([]*TabletHealth, 0, len(all))
for _, s := range all {
// only tablets in same cell / cellAlias are included in healthy list
if hc.isIncluded(shr) {
// Only tablets in same cell / cellAlias are included in healthy list.
if hc.isIncluded(s.Tablet.Type, s.Tablet.Alias) {
allArray = append(allArray, s)
}
}
Expand Down Expand Up @@ -686,14 +689,14 @@ func (hc *HealthCheckImpl) getAliasByCell(cell string) string {
return alias
}

func (hc *HealthCheckImpl) isIncluded(shr *query.StreamHealthResponse) bool {
if shr.Target.TabletType == topodata.TabletType_MASTER {
func (hc *HealthCheckImpl) isIncluded(tabletType topodata.TabletType, tabletAlias *topodata.TabletAlias) bool {
if tabletType == topodata.TabletType_MASTER {
return true
}
if shr.TabletAlias.Cell == hc.cell {
if tabletAlias.Cell == hc.cell {
return true
}
if hc.getAliasByCell(shr.TabletAlias.Cell) == hc.getAliasByCell(hc.cell) {
if hc.getAliasByCell(tabletAlias.Cell) == hc.getAliasByCell(hc.cell) {
return true
}
return false
Expand Down