Skip to content
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

Jump to bottom

Don't start a learner node with no peers #7582

Merged
merged 2 commits into from
Mar 17, 2021
Merged

Don't start a learner node with no peers #7582

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1c680a1
Don't start a learner node with no peers
ahsanbarkati Mar 15, 2021
96d298a
Add a todo
ahsanbarkati Mar 16, 2021
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
8 changes: 7 additions & 1 deletion worker/draft.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1729,7 +1729,8 @@ func (n *node) InitAndStartNode() {
_, restart, err := n.PastLife()
x.Check(err)

if _, hasPeer := groups().MyPeer(); !restart && hasPeer {
_, hasPeer := groups().MyPeer()
if !restart && hasPeer {
// The node has other peers, it might have crashed after joining the cluster and before
// writing a snapshot. Check from leader, if it is part of the cluster. Consider this a
// restart if it is part of the cluster, else start a new node.
Expand All @@ -1742,6 +1743,10 @@ func (n *node) InitAndStartNode() {
}
}

if n.RaftContext.IsLearner && !hasPeer {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For the case of restart, should we check again after making connections to nodes in ConfState? I don't know why it is required.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I suggest we keep this here for now.

glog.Fatal("Cannot start a learner node without peer alpha nodes")
}

if restart {
glog.Infof("Restarting node for group: %d\n", n.gid)
sp, err := n.Store.Snapshot()
Expand All @@ -1754,6 +1759,7 @@ func (n *node) InitAndStartNode() {
// zero-member Raft group.
n.SetConfState(&sp.Metadata.ConfState)

// TODO: Making connections here seems unnecessary, evaluate.
members := groups().members(n.gid)
for _, id := range sp.Metadata.ConfState.Nodes {
m, ok := members[id]
Expand Down