The replicant problem of historical node

Hi guys:
I have used druid for a period time. In the beginning, the druid cluster had only one historical node.

Recently, I run another historical node( the configure is the same as the first one) . But the second node

loaded the same segment cache as the first one. Does durid have replicant segment cache when

run multi historical node? If so, what is the default repilcant number?

You should take a look at your coordinator console:

You can configure your replication rules !

Hi Benjamin:
****I haven’t configure any replication rules in coordinator console web page, and in coordinator console page the replicant is 1.

Is the default replicant of druid is 2?

If the default replicant is 1 on your coordinator console, you should have only 1 copy of each segment.
Are the historical nodes on the same tiers?

How have you created your second historical?

Hi Benjamin:
the two historical nodes are on the same tiers, the _default_tier.

I copied the config file from first historical node, only changed the service name. In fact, these two nodes, the configuration are nearly the same.

The config is in the attachment, is there anything wrong?

在 2016年8月24日星期三 UTC+8下午3:47:35,Benjamin Angelaud写道:

Druid will try to store 2 copies of any segment in a cluster. As Ben mentioned, you can disable this behaviour with rules. If you only have 2 historicals, they will be exact copies of each other.

Thanks, I got it!

在 2016年8月26日星期五 UTC+8上午5:58:39,Fangjin Yang写道: