You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@cassandra.apache.org by Andrey Ilinykh <ai...@gmail.com> on 2013/12/05 20:31:59 UTC

vnodes on aws

Hello everybody!
We run cassandra 1.1 on ec2 instances. We use three availability zones, the
replication factor is 3 also. NetworkTopologyStrategy guarantees each row
is replicated in all availability zones. So, if we lost one zone quorum
operations still work. We think about to upgrade to 1.2. Virtual nodes are
the main reason. My understanding is - vnodes are distributed randomly, so
their is no way to put every row into all availability zones.  Am I right?
What would be the best way to deploy vnodes across several data centers
(availability zones)?


Thank you,
  Andrey

Re: vnodes on aws

Posted by Robert Coli <rc...@eventbrite.com>.
On Thu, Dec 5, 2013 at 6:58 PM, Andrey Ilinykh <ai...@gmail.com> wrote:

>
>
>
> On Thu, Dec 5, 2013 at 3:31 PM, Jayadev Jayaraman <jd...@gmail.com>wrote:
>
>> Availability zones are analogous to racks not data centres . EC2 regions
>> are equivalent to data centres.
>>
> Yes, this is what I meant. I guess my question is - is possible to put row
> in every rack using vnodes?
>

https://issues.apache.org/jira/browse/CASSANDRA-4123
"
I should point out that the existing NTS (and also the modified
implementation in CASSANDRA-3881) already prevent replicas from being
placed on the same host. SimpleStrategy also is fixed in the CASSANDRA-4121
patch. The first couple of paragraphs in the ticket description could
probably be removed.
"

https://issues.apache.org/jira/browse/CASSANDRA-4121
https://issues.apache.org/jira/browse/CASSANDRA-3881

=Rob

Re: vnodes on aws

Posted by Andrey Ilinykh <ai...@gmail.com>.
On Thu, Dec 5, 2013 at 3:31 PM, Jayadev Jayaraman <jd...@gmail.com>wrote:

> Availability zones are analogous to racks not data centres . EC2 regions
> are equivalent to data centres.
>
Yes, this is what I meant. I guess my question is - is possible to put row
in every rack using vnodes?

Thank you,
   Andrey



> On Dec 5, 2013 2:32 PM, "Andrey Ilinykh" <ai...@gmail.com> wrote:
>
>> Hello everybody!
>> We run cassandra 1.1 on ec2 instances. We use three availability zones,
>> the replication factor is 3 also. NetworkTopologyStrategy guarantees each
>> row is replicated in all availability zones. So, if we lost one zone quorum
>> operations still work. We think about to upgrade to 1.2. Virtual nodes are
>> the main reason. My understanding is - vnodes are distributed randomly, so
>> their is no way to put every row into all availability zones.  Am I right?
>> What would be the best way to deploy vnodes across several data centers
>> (availability zones)?
>>
>>
>> Thank you,
>>   Andrey
>>
>

Re: vnodes on aws

Posted by Jayadev Jayaraman <jd...@gmail.com>.
Availability zones are analogous to racks not data centres . EC2 regions
are equivalent to data centres.

You can use vnodes if you want to deploy a cluster across multiple regions
(data centres) with one availability zone per region. Each region maintains
a separate ring.

I don't know if you can use vnodes on multiple rack topologies however. In
my case, Wlwe host a EC2 cluster with 2 fold replication and 2 zones in the
same region and we don't use vnodes
On Dec 5, 2013 2:32 PM, "Andrey Ilinykh" <ai...@gmail.com> wrote:

> Hello everybody!
> We run cassandra 1.1 on ec2 instances. We use three availability zones,
> the replication factor is 3 also. NetworkTopologyStrategy guarantees each
> row is replicated in all availability zones. So, if we lost one zone quorum
> operations still work. We think about to upgrade to 1.2. Virtual nodes are
> the main reason. My understanding is - vnodes are distributed randomly, so
> their is no way to put every row into all availability zones.  Am I right?
> What would be the best way to deploy vnodes across several data centers
> (availability zones)?
>
>
> Thank you,
>   Andrey
>