Skip to content

KeepLeadership can swallow exceptions #133

@AzureView

Description

@AzureView

If using the leader function we find that there is a problem with the failover.

If have a single instance of service which has aquired the leadership lock, but then kill consul on that server. When subsequently we restart consul the service should reaquire the leadership lock (assuming there was no other instance of the service running which had aquired the lock). However it does not re-aquire the leadership lock.

Similary if have a second instance of the service running which aquires the lock when consul is killed on the server where the instance had originally aquired the lock - then thre is no log message to say the original instance lost the lock.

I believe the issue that needs to be fixe is that any exception in the KeepLeadership loop in LeaderWatcher is swallower and the loop stops. It may need to have try catch similar tot he the watcherLoop in ServiceWatcher.

Thanks Derek

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions