Shutdown and start order in a clustered /distributed architecture

[giorgiovolpini]

HI,

I have an environment with the following architecture: 2 Search Heads, 2 Idx, 1 Dep Server, 1 Cluster Master, 2 Heavy forwarder, >500 UF.

I have to shoot down and than start all the nodes.

My question is: does exists a specific order or precedure to follow? What is the best practice in this cases?

Thank you

Giorgio

[gcusello]

HI @giorgiovolpini,
for the Splunk Servers, the only behaviour is to not shutdown at the same time both the servers of the same kind.
For Universal forwarders, there isn't any rule.

In othe words, there isn't an order to shutdown: shutdown all the servers one by one.
It's different if you have to upgrade Splunk, but it's a different question!

Ciao.
Giuseppe