I couldn’t find anywhere any answer how to run DDL on production cluster with multiple schema.
In 3 nodes production cluster I have about 10 schema, each one is dedicated for different Client. Databases are rather big, there are tables which have over 7mln of records.
From time to time there is a need to alter table. Unfortunately wherever I have to make something like that not only one schema is not available but all databases All environment is being blocked by DDL operation which is dramatic and has nothing common with HA idea.
My temporary solution which I don’t like:
I disconnect one node from the cluster, run DDL on this disconnected node, then I make a dump and recovery it on running DB Cluster. After all this steps I am preparing disconnected node to connect to cluster with SST.
During this all horrible steps all databases are on-line which is the most important for me.
What I would like to have:
Please is there any other better solution to run DDL without affecting any others DATABASES except this one on each DDL is running ?