incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From aaron morton <aa...@thelastpickle.com>
Subject Re: nodetool status OWNS and multiple DCs
Date Tue, 30 Apr 2013 20:11:06 GMT
> I thought that each datacenter has 100% coverage of token range. What does the value in
"Owns" field mean and how it affects a replication (for exaple, with replication factors DC1:1,
DC2:2)?
Run the command and specify your keyspace, that will tell nodetool to use the Replication
Strategy specified for KS when calculating the layout. If it's NTS you should see what you
expect. 

Cheers
  
-----------------
Aaron Morton
Freelance Cassandra Consultant
New Zealand

@aaronmorton
http://www.thelastpickle.com

On 30/04/2013, at 8:50 PM, Sergey Naumov <sknaumov@gmail.com> wrote:

> Hello.
> 
> I have set up test cluster of 2 DCs with 1 node in each DC. In each config I specified
256 virtual nodes and chosed GossippingPropertyFileSnitch.
> 
> For node1:
> ~/Cassandra$ cat /etc/cassandra/cassandra-rackdc.properties
> dc=DC1
> rack=RAC1
> 
> For node2:
> ~/Cassandra$ cat /etc/cassandra/cassandra-rackdc.properties
> dc=DC2
> rack=RAC1
> 
> When I call "nodetool status", it shows not 100% ownership of tokens for each DC:
> :~/Cassandra$ nodetool status
> Datacenter: DC1
> ===============
> Status=Up/Down
> |/ State=Normal/Leaving/Joining/Moving
> --  Address     Load       Tokens  Owns   Host ID                               Rack
> UN  172.16.0.1  97.43 KB   256     51,5%  7ffd2432-46c9-443c-aa0c-8bfd960d2acc  RAC1
> Datacenter: DC2
> ===============
> Status=Up/Down
> |/ State=Normal/Leaving/Joining/Moving
> --  Address     Load       Tokens  Owns   Host ID                               Rack
> UN  172.16.0.2  91.78 KB   256     48,5%  93d3bdbf-8625-4e54-8c7c-087fbfe419f5  RAC1
> 
> I thought that each datacenter has 100% coverage of token range. What does the value
in "Owns" field mean and how it affects a replication (for exaple, with replication factors
DC1:1, DC2:2)?
> 
> Thanks in advance,
> Sergey Naumov.


Mime
View raw message