Support Forums : DataStax

DataStax Support Forums » Apache Cassandra

why I have to setup Node IP in "seeds" list to be able run local on port 9160?

(3 posts) (2 voices)

Started 6 months ago by mikek753
Latest reply from mikek753
This topic is resolved

Tags:

No tags yet.

mikek753
Member

Hello experts,

I have some issue or misunderstanding about Cassandra settings in cluster mode.
I need to know why I have to setup Node IP in "seeds" list to be able run local on port 9160? otherwise local access doesn't work!!!

Based on Cassandra doc the seeds list has to be exact the same for all Nodes.
However, any new Node will join with predefined list, but 127.0.0.1:9160 will not work until new Node interface (Eth0) in local seeds list. Cassandra doesn’t bind to 9160 at all.
There is no way I want to update other existing Nodes to have any new Node IP in own seeds list – this is when Cluster has above 5 Nodes, just isn’t practical.

How to solve this settings issue?
I'd like to add "power on" and later "power off" Cassandra Nodes on demand, while keep 2-5 Nodes ON all the time and use those for bootstrap for other Nodes

Without own IP in seeds list Node would join the Cluster, but will not bind to local host 9160.
Yes, I have for rpc_address 0.0.0.0

Thanks for guidance.

Posted 6 months ago #
nickmbailey
Key Master

I'm not sure if I understand your problem. You want thrift to bind to all interfaces on the machine but it will only bind to all interfaces when you put its own ip in the seed list?

Your problem may be that a node will not bootstrap when it is in the seed list, so by putting it there you are preventing bootstrapping. Thrift will not start up until bootstrapping is complete, so you need to wait for that to finish before checking to see if thrift is started.

Posted 6 months ago #
mikek753
Member

thank you for the reply.
Yes, when I removed own IP from seeds list it took about a min, but thrift started and see in the log that Cassandra binds to 9160.

May be this delay can be very long to get local client worry.
I'll try to recreate cluster to see how long it takes to bootstrap initially.

What I have - 5 nodes cluster.
What I did on secondary node
1. stop Cassandra
2. removed own IP from seeds list
3. removed data folder
4. start Cassandra

in the log I saw that another Node contacted this and I didn't see bootstrap from "primary" (only IP in seeds) node. And after that another nodes appears.

So far so good.

Posted 6 months ago #

RSS feed for this topic

Reply

You must log in to post.