You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When adding a node to a cluster, you must specify the address of at least one existing node in the --join flag. We should update production deployment docs to recommend putting multiple node addresses in there, so the node can restart even if one of the nodes in join isn't available.
In local startup docs, probably ok to leave the code samples with one other address in join, but we could mention this in the supporting text.
Also, we should clarify that --join is not required on node restart. The only exception is if, when a node was started, it was shut down before it saved its peer address to disk.
jseldess
changed the title
Production recommendation: include all node addresses in --join
Production recommendation: include multiple node addresses in --join
Oct 10, 2016
When adding a node to a cluster, you must specify the address of at least one existing node in the
--join
flag. We should update production deployment docs to recommend putting multiple node addresses in there, so the node can restart even if one of the nodes injoin
isn't available.In local startup docs, probably ok to leave the code samples with one other address in
join
, but we could mention this in the supporting text.Also, we should clarify that
--join
is not required on node restart. The only exception is if, when a node was started, it was shut down before it saved its peer address to disk.@knz, @aphyr
The text was updated successfully, but these errors were encountered: