cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ananthkumar K S (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-7861) Node is not able to gossip
Date Thu, 04 Sep 2014 05:33:51 GMT


Ananthkumar K S commented on CASSANDRA-7861:

[~brandon.williams] No. These two nodes were set up nearly four months back. We restarted
the two nodes post taking the above mentioned thread dump. These two nodes were denoted as
DN from other data center . But we were able to login to the server and see the cassandra
nodes running in this. If we do a nodetool status in this , we get DN for nodes in other data
center . When we checked the logs, OutboundTCPConnection was failing to all the nodes in other
data center. 

If TCP connection is a problem , how does a node restart solve this issue? So network is up
and running. Its true that we had a slight connection issue for , say 30 mins, where the network
was really jittery. But cassandra is designed to handle these kind of problems right?

> Node is not able to gossip
> --------------------------
>                 Key: CASSANDRA-7861
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Ananthkumar K S
>             Fix For: 2.0.3
>         Attachments: thread-dump.txt
> The node is running on All of a sudden, it was not able to gossip and
find the other nodes between data centres. We had two nodes indicated as down in DC1 but those
two nodes were up and running in DC2. When we check the two nodes status in DC2, all the nodes
in DC1 are denoted as DN and the other node in DC2 is denoted as down. 
> There seems to be a disconnect between the nodes. I have attached the thread dump of
the node that was down. 

This message was sent by Atlassian JIRA

View raw message