NoNodeAvailableException: No node was available to execute the query

Question

I am not using Elasticssearch. I am trying to perform some database operations in cassandra using CQL. I am using threads. While running the code I am always getting the exception in thread after a while : com.datastax.oss.driver.api.core.NoNodeAvailableException: No node was available to execute the query.

I have tested with even one thread. The error is still there. Here is my code :

InetAddress addrOne = InetAddress.getByName("52.15.195.41");
InetSocketAddress addrSocOne = new InetSocketAddress(addrOne,9042);
CqlSession sessionOne = CqlSession.builder().addContactPoint(addrSocOne).withLocalDatacenter("us-east-2").withKeyspace("test").build();

while(counter <= 100)
{
    String query = "select max(id) FROM samplequeue";
    ResultSet rs = session.execute(query);
    for (Row row : rs) 
    {
        int exS = row.getInt("system.max(id)");
    }
    counter++;
    Thread.sleep(50);
}

This is a very simple, modified example just to demonstrate the problem. I am unable to resolve it. All the threads are exiting giving the same exception. I am running cassandra 3.11.4 on AWS. All my nodes are up and running and I can perform operations finely in the backend.

@CarlosMonroyNieblas I have added that in the code section of my question now. — SDt, Oct 08 '19 at 05:41
William brings up a good point below. When you run `nodetool status` what is the name of your data center? — Aaron, Oct 29 '19 at 12:31

score 17 · Accepted Answer · edited Oct 28 '19 at 10:23

17

Change .withLocalDatacenter("us-east-2") to .withLocalDatacenter("datacenter1") and retry.

edited Oct 28 '19 at 10:23

JoSSte

2,953
6
34
54

answered Oct 28 '19 at 09:45

William

194
1
3

7

could you please explain why this works? is it some hard coded value? – Leonid Bor Nov 17 '19 at 16:15
I'm surprised that this worked, but it did, even though my datacenter's name is different than "datacenter1" (as confirmed by `nodetool status`). Setting to "datacenter1" works though. For me, before I did this I would get inconsistent results, sometimes working sometimes not, but after I did this, it worked great. – RyanQuey Oct 14 '20 at 03:52
Update: it seems like it might relate to whether or not there is an instance of C* running on the machine using the Java driver. I was connecting to C* running on an external AWS, but had C* (specifically DSE) running on my local box, and had to use .withLocalDatacenter("datacenter1"). But when I stopped the local C* instance (sudo service dse stop), then tried again, now it couldn't connect, and complained giving the error as shown in next comment – RyanQuey Oct 14 '20 at 04:44
`You specified datacenter1 as the local DC, but some contact points ar e from a different DC: Node(endPoint=/:9042, hostId=, hashCode=4a395efc)=SearchGraph; please provide the correct local DC, or check your contact points` – RyanQuey Oct 14 '20 at 04:44

score 15 · Answer 2 · edited Feb 20 '20 at 10:37

15

You can execute nodetool status in cassandra and you will obtain the name of the Datacenter. That is the value that you must put in withLocalDatacenter method:

/opt/apache-cassandra-3.11.2/bin$ ./nodetool status
Datacenter: **dc1**
===============
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address         Load       Tokens       Owns (effective)  Host ID    Rack

edited Feb 20 '20 at 10:37

B--rian

5,578
10
38
89

answered Feb 20 '20 at 10:16

Ceci G J

151
1
2

This solution worked for me. In my case, I was using wrong data centre name DC1 in my application. But it was datacentre1 which I found with this "nodetools status" command. – Gunjan Shah Jul 27 '22 at 14:33

score 0 · Answer 3 · answered Nov 29 '21 at 14:04

0

When running in local, better to run nodetool status command get the dc name and use in application while connecting to Cassandra instance.

answered Nov 29 '21 at 14:04

PraveenKumar K G

176
1
7

NoNodeAvailableException: No node was available to execute the query

3 Answers3