Doubt in Host.StateListener

13 views
Skip to first unread message

Sumit Thakur

unread,
May 20, 2015, 8:01:12 AM5/20/15
to java-dri...@lists.datastax.com
Hello Team,

I am getting wrong message  while trying to test state listener.

step 1: cassandra - 10.49.16.73 is killed.
output :   onDown : Event [eventType=onDown, host=/10.49.16.73:9042]

step 2.  cassandra - 10.49.16.73 is up. 
output:  
            onUp : Event [eventType=onUp, host=/10.49.16.73:9042]
            onDown : Event [eventType=onDown, host=/10.49.16.73:9042]
            onUp : Event [eventType=onUp, host=/10.49.16.73:9042]
            onDown : Event [eventType=onDown, host=/10.49.16.73:9042 

1. why i am getting 4 message (above log ) in cassandra restart, (up,down,up,down).
Note: client is also running in 10.49.16.73.  

2. if  remote cassandra is killed ,then first goes in suspected and then down state. (why?)

3. In jconsole , i saw RMI TCP Connection(3)-10.49.16.73  :: ( RMI TCP connection is only for  local running cassandra, why not for remote cassandra ??? )
    


Thanks & Regards
Sumit Thakur


Sumit Thakur

unread,
May 22, 2015, 7:56:44 AM5/22/15
to java-dri...@lists.datastax.com
Still waiting for reply .........

Kevin Gallardo

unread,
May 22, 2015, 11:29:24 AM5/22/15
to java-dri...@lists.datastax.com
Hello, 

I think the best thing to do first would be to give us your driver version and the version of Cassandra you are using it on, so we can have a better idea of the range of issues you can run into.
The suspected state has been deprecated since the last version of the driver, but the behaviour you are describing here seems quite fair. As described in the Documentation, when an error happens on one connection of a node (like a timeout because the "remote cassandra" is killed), this node is set Suspected, then after a failed reconnection attempt, the node is set Down.
Reply all
Reply to author
Forward
0 new messages