Closed phlogistonjohn closed 2 months ago
I was shown a cluster that had failed to communicate with rados during the ctdb-must-have-node command, after it had at least correctly read "nothing". Add a retry loop (with limit) to the command to make it robust against spurious errors.
ctdb-must-have-node
@mergifyio rebase
rebase
@mergifyio refresh
refresh
ok mergify whatever.
I was shown a cluster that had failed to communicate with rados during the
ctdb-must-have-node
command, after it had at least correctly read "nothing". Add a retry loop (with limit) to the command to make it robust against spurious errors.