Unsafe cluster bootstrapping is, as the name suggests, unsafe. It risks silently losing some of your data so you cannot be sure that the data did not change; worse, Elasticsearch has no way to identify any lost data. Seems like you were lucky this time, but there's no replacement for setting up (and testing!) a proper snapshot and restore process.
3 Likes
Indeed, you are right. This time is a big lesson for me. I also realize how important the daily maintenance of a cluster, correct configuration and data backup are! Very appreciate it for you.
But I think it is really important that you can provide Unsafe cluster bootstrapping. Because of the differences in the knowledge level of operation and maintenance personnel and the particularity of various deployment environments, it is easy for a situation like mine to occur, and I was almost there. I just watched those indices files but couldn't restore them, almost starting from scratch. 
I hope you will never ban or remove this Unsafe cluster bootstrapping in the future. 
We are talking about adding checks when you start up a new cluster to prevent this type of situations, not to remove or alter the elasticsearch-node tool.
1 Like
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.