Wiki contributions

[a.porfirio]

Hello,
I would like to contribute to the Apache Cassandra wiki: my wiki username is 
AlicePorfirio.

I already tried to post this request, but I am not sure if that email was 
correctly sent: in case it was, excuse me for the duplicate.

Thank you very much.

Best regards,
Alice Porfirio

Re: Network transfer to one node twice as others

[Benedict Elliott Smith]

If you're connecting via thrift, all your traffic is most likely being
routed to just one node, which then communicates with the other nodes for
you.

On Wed, Apr 22, 2015 at 6:11 AM, Anishek Agarwal  wrote:

> Forwarding it here, someone with Cassandra internals knowledge can help may
> be
>
> Additionally, i observe the same behavior for reads too where Network read
> from one node is twice than other two..
>
>
> -- Forwarded message --
> From: Anishek Agarwal 
> Date: Tue, Apr 21, 2015 at 5:15 PM
> Subject: Network transfer to one node twice as others
> To: "u...@cassandra.apache.org" 
>
>
> Hello,
>
> We are using cassandra 2.0.14 and have a cluster of 3 nodes. I have a
> writer test (written in java) that runs 50 threads to populate data to a
> single table in a single keyspace.
>
> when i look at the "iftop"  I see that the amount of network transfer
> happening on two nodes is same but on one of the nodes its almost 2ice as
> the other two, Any reason that would be the case ?
>
> Thanks
> Anishek
>

Re: Network transfer to one node twice as others

[Anishek Agarwal]

Nope not using thrift
On 22-Apr-2015 7:24 pm, "Benedict Elliott Smith" 
wrote:

> If you're connecting via thrift, all your traffic is most likely being
> routed to just one node, which then communicates with the other nodes for
> you.
>
> On Wed, Apr 22, 2015 at 6:11 AM, Anishek Agarwal 
> wrote:
>
> > Forwarding it here, someone with Cassandra internals knowledge can help
> may
> > be
> >
> > Additionally, i observe the same behavior for reads too where Network
> read
> > from one node is twice than other two..
> >
> >
> > -- Forwarded message --
> > From: Anishek Agarwal 
> > Date: Tue, Apr 21, 2015 at 5:15 PM
> > Subject: Network transfer to one node twice as others
> > To: "u...@cassandra.apache.org" 
> >
> >
> > Hello,
> >
> > We are using cassandra 2.0.14 and have a cluster of 3 nodes. I have a
> > writer test (written in java) that runs 50 threads to populate data to a
> > single table in a single keyspace.
> >
> > when i look at the "iftop"  I see that the amount of network transfer
> > happening on two nodes is same but on one of the nodes its almost 2ice as
> > the other two, Any reason that would be the case ?
> >
> > Thanks
> > Anishek
> >
>

Re: Network transfer to one node twice as others

[Ryan Svihla]

Totally depends on the load balancing policy of your driver, your data model, 
consistently level and you’re replication factor. The default token aware 
policy for the DataStax java driver up to 2.1.4 and 2.0.9 would largely behave 
this way if you combined it with a hot partition, and all other languages of 
the DataStax driver to my knowledge behave the same way today.

If you’re using the DataStax driver go ahead and change the load balancing 
policy to something like DCAwareRoundRobin only and see if the result is 
different. 

Honestly this is probably not a question for the dev mailing list, or even the 
Cassandra user mailing list but likely for whatever driver you’re using.

- Ryan

> On Apr 22, 2015, at 9:16 AM, Anishek Agarwal  wrote:
> 
> Nope not using thrift
> On 22-Apr-2015 7:24 pm, "Benedict Elliott Smith" 
> wrote:
> 
>> If you're connecting via thrift, all your traffic is most likely being
>> routed to just one node, which then communicates with the other nodes for
>> you.
>> 
>> On Wed, Apr 22, 2015 at 6:11 AM, Anishek Agarwal 
>> wrote:
>> 
>>> Forwarding it here, someone with Cassandra internals knowledge can help
>> may
>>> be
>>> 
>>> Additionally, i observe the same behavior for reads too where Network
>> read
>>> from one node is twice than other two..
>>> 
>>> 
>>> -- Forwarded message --
>>> From: Anishek Agarwal 
>>> Date: Tue, Apr 21, 2015 at 5:15 PM
>>> Subject: Network transfer to one node twice as others
>>> To: "u...@cassandra.apache.org" 
>>> 
>>> 
>>> Hello,
>>> 
>>> We are using cassandra 2.0.14 and have a cluster of 3 nodes. I have a
>>> writer test (written in java) that runs 50 threads to populate data to a
>>> single table in a single keyspace.
>>> 
>>> when i look at the "iftop"  I see that the amount of network transfer
>>> happening on two nodes is same but on one of the nodes its almost 2ice as
>>> the other two, Any reason that would be the case ?
>>> 
>>> Thanks
>>> Anishek
>>> 
>> 

Re: Wiki contributions

[Brandon Williams]

I have added you.

On Wed, Apr 22, 2015 at 4:34 AM,  wrote:

> Hello,
> I would like to contribute to the Apache Cassandra wiki: my wiki username
> is AlicePorfirio.
>
> I already tried to post this request, but I am not sure if that email was
> correctly sent: in case it was, excuse me for the duplicate.
>
> Thank you very much.
>
> Best regards,
> Alice Porfirio
>
>

A proposal for how we use JIRA in the tick-tock release process

[Ryan McGuire]

In the interests of making the tick tock release process as smooth and

efficient as possible, I’d like to propose a few procedural JIRA

rules:


 * Let’s use the In Progress status to indicate when development is

actually in progress. This can be a very useful indicator to testers

that it’s the right time to engage the developer to discuss testing

plans and agree on the Definition of Done for that ticket.


 * Let’s use the Testing status after a patch has been reviewed, and

before the patch gets merged, to be an opportunity for people to chime

in about whether or not the proposed change has adequate testing and

meets the Definition of Done.


It’s not my intention to add needless formalities to the process or to

slow things down for the developers - test planning and test

implementation should always be done concurrently while a test is In

Progress, so the Testing status for a ticket should be short lived.

What it gives us is a more solid way of knowing that what gets merged

into trunk is in as best shape as it can be, and is always

deliverable.


I would also note that the Testing phase should not be regarded as

only for the DataStax test engineering team. It really should be a

collaborative phase where we all can discuss the tests that everyone

has contributed. If a developer is confident that all the testing is

in place (unit tests, dtests, etc.) then they should feel free to skip

the testing status.


--

[image: datastax_logo.png] 

Ryan McGuire

Software Engineering Manager in Test | r...@datastax.com

[image: linkedin.png]  [image:
twitter.png] 

Re: A proposal for how we use JIRA in the tick-tock release process

[Jonathan Ellis]

SGTM.

On Thu, Apr 23, 2015 at 1:48 AM, Ryan McGuire  wrote:

> In the interests of making the tick tock release process as smooth and
>
> efficient as possible, I’d like to propose a few procedural JIRA
>
> rules:
>
>
>  * Let’s use the In Progress status to indicate when development is
>
> actually in progress. This can be a very useful indicator to testers
>
> that it’s the right time to engage the developer to discuss testing
>
> plans and agree on the Definition of Done for that ticket.
>
>
>  * Let’s use the Testing status after a patch has been reviewed, and
>
> before the patch gets merged, to be an opportunity for people to chime
>
> in about whether or not the proposed change has adequate testing and
>
> meets the Definition of Done.
>
>
> It’s not my intention to add needless formalities to the process or to
>
> slow things down for the developers - test planning and test
>
> implementation should always be done concurrently while a test is In
>
> Progress, so the Testing status for a ticket should be short lived.
>
> What it gives us is a more solid way of knowing that what gets merged
>
> into trunk is in as best shape as it can be, and is always
>
> deliverable.
>
>
> I would also note that the Testing phase should not be regarded as
>
> only for the DataStax test engineering team. It really should be a
>
> collaborative phase where we all can discuss the tests that everyone
>
> has contributed. If a developer is confident that all the testing is
>
> in place (unit tests, dtests, etc.) then they should feel free to skip
>
> the testing status.
>
>
> --
>
> [image: datastax_logo.png] 
>
> Ryan McGuire
>
> Software Engineering Manager in Test | r...@datastax.com
>
> [image: linkedin.png]  [image:
> twitter.png] 
> 
>



-- 
Jonathan Ellis
Project Chair, Apache Cassandra
co-founder, http://www.datastax.com
@spyced

Re: Network transfer to one node twice as others

[Anishek Agarwal]

Thanks Ryan,

I am using the java driver with token aware policy(dcawarepolicy) we have
all nodes on one DC. model for now is
 partition_key, value

I think by hot partition you mean insertions to the same partition key ? we
have RF:3 and CL: 1 (will re run with local quorum as that is what we will
use eventually)

Closing this discussion here, reposting it on javax driver mailing list.




On Wed, Apr 22, 2015 at 7:56 PM, Ryan Svihla  wrote:

> Totally depends on the load balancing policy of your driver, your data
> model, consistently level and you’re replication factor. The default token
> aware policy for the DataStax java driver up to 2.1.4 and 2.0.9 would
> largely behave this way if you combined it with a hot partition, and all
> other languages of the DataStax driver to my knowledge behave the same way
> today.
>
> If you’re using the DataStax driver go ahead and change the load balancing
> policy to something like DCAwareRoundRobin only and see if the result is
> different.
>
> Honestly this is probably not a question for the dev mailing list, or even
> the Cassandra user mailing list but likely for whatever driver you’re using.
>
> - Ryan
>
> > On Apr 22, 2015, at 9:16 AM, Anishek Agarwal  wrote:
> >
> > Nope not using thrift
> > On 22-Apr-2015 7:24 pm, "Benedict Elliott Smith" <
> belliottsm...@datastax.com>
> > wrote:
> >
> >> If you're connecting via thrift, all your traffic is most likely being
> >> routed to just one node, which then communicates with the other nodes
> for
> >> you.
> >>
> >> On Wed, Apr 22, 2015 at 6:11 AM, Anishek Agarwal 
> >> wrote:
> >>
> >>> Forwarding it here, someone with Cassandra internals knowledge can help
> >> may
> >>> be
> >>>
> >>> Additionally, i observe the same behavior for reads too where Network
> >> read
> >>> from one node is twice than other two..
> >>>
> >>>
> >>> -- Forwarded message --
> >>> From: Anishek Agarwal 
> >>> Date: Tue, Apr 21, 2015 at 5:15 PM
> >>> Subject: Network transfer to one node twice as others
> >>> To: "u...@cassandra.apache.org" 
> >>>
> >>>
> >>> Hello,
> >>>
> >>> We are using cassandra 2.0.14 and have a cluster of 3 nodes. I have a
> >>> writer test (written in java) that runs 50 threads to populate data to
> a
> >>> single table in a single keyspace.
> >>>
> >>> when i look at the "iftop"  I see that the amount of network transfer
> >>> happening on two nodes is same but on one of the nodes its almost 2ice
> as
> >>> the other two, Any reason that would be the case ?
> >>>
> >>> Thanks
> >>> Anishek
> >>>
> >>
>
>