You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@chukwa.apache.org by William Bajzek <wi...@gmail.com> on 2010/08/03 23:26:33 UTC

Re: What to do after database failures?

Hi Eric, 

I'm not sure how many people are really using MySQL in this capacity. I know that one of the other people who has posted to the mailing list about i recently is my coworker, so it's possible that our recent activity is giving the impression of more demand than there really is. Our client is a big corporation who plans to use Chukwa this way and I'm working on a patch for them that does as you say. It is pretty small and straightforward. What would be a good way to contribute it back to the community? 



On Jul 23, 2010, at 9:58 AM, Eric Yang wrote:
> 
> MetricDataLoader can be modified to throw IOException to the executor class
> MetricDataLoaderPool, which can throw exception to PostProcessor Manager.
> PostProcessorManager moves the data to a temp directory.  The retry logic
> can be added to PostProcessorManager by counting the number of retry with
> the error out sequence file before sending it to InError directory.  It
> should be the better route to manages error conditions.
> 
> It seems like a lot of people are using MySQL part of Chukwa.  I am
> wondering what are the use cases?  If we can have a driver on this part of
> the project, it could be better supported without fragment the community.



- William Bajzek
williambajzek@gmail.com





Re: What to do after database failures?

Posted by Ariel Rabkin <as...@gmail.com>.
Hrm.  They were working for me a few days ago.  Go ahead and upload
the patch anyway.  It's okay to post a patch slightly flakey code.

--Ari

On Tue, Aug 3, 2010 at 3:59 PM, William Bajzek <wi...@gmail.com> wrote:
> Hi Eric & Ari,
>
> Thanks for the info. Right now I can't get the unit tests to pass even with a freshly checked out (without my code) copy from trunk OR the 4.0 branch, so I will hold off on submitting my patch until I can get that figured out.
>
> - William Bajzek
> williambajzek@gmail.com
>
>
>
>
>



-- 
Ari Rabkin asrabkin@gmail.com
UC Berkeley Computer Science Department

Re: What to do after database failures?

Posted by William Bajzek <wi...@gmail.com>.
Hi Eric & Ari,

Thanks for the info. Right now I can't get the unit tests to pass even with a freshly checked out (without my code) copy from trunk OR the 4.0 branch, so I will hold off on submitting my patch until I can get that figured out. 

- William Bajzek
williambajzek@gmail.com





Re: What to do after database failures?

Posted by Ariel Rabkin <as...@gmail.com>.
The right way to contribute it back is to open a JIRA, describe what
the patch does, and attach it as a file.
We'll try to review and commit it within a couple days, barring problems.

--Ari

On Tue, Aug 3, 2010 at 2:26 PM, William Bajzek <wi...@gmail.com> wrote:
> Hi Eric,
>
> I'm not sure how many people are really using MySQL in this capacity. I know that one of the other people who has posted to the mailing list about i recently is my coworker, so it's possible that our recent activity is giving the impression of more demand than there really is. Our client is a big corporation who plans to use Chukwa this way and I'm working on a patch for them that does as you say. It is pretty small and straightforward. What would be a good way to contribute it back to the community?
>
>
>
> On Jul 23, 2010, at 9:58 AM, Eric Yang wrote:
>>
>> MetricDataLoader can be modified to throw IOException to the executor class
>> MetricDataLoaderPool, which can throw exception to PostProcessor Manager.
>> PostProcessorManager moves the data to a temp directory.  The retry logic
>> can be added to PostProcessorManager by counting the number of retry with
>> the error out sequence file before sending it to InError directory.  It
>> should be the better route to manages error conditions.
>>
>> It seems like a lot of people are using MySQL part of Chukwa.  I am
>> wondering what are the use cases?  If we can have a driver on this part of
>> the project, it could be better supported without fragment the community.
>
>
>
> - William Bajzek
> williambajzek@gmail.com
>
>
>
>
>



-- 
Ari Rabkin asrabkin@gmail.com
UC Berkeley Computer Science Department

Re: What to do after database failures?

Posted by Eric Yang <ey...@yahoo-inc.com>.
Hi William,

The best approach is to submit patches.  Chukwa has a bug tracking system
which includes all development activities.
See this page:

https://issues.apache.org/jira/browse/CHUKWA

Patches are reviewed by the community to make sure it works and license
compatible.  Once a active individual contributor has been submitting
patches for over a period of time, he/she is promoted to become a committer.
A committer will have svn access to commit code.  However, all changes still
go through the same review process as before.  Hope this helps.

Regards,
Eric

On 8/3/10 2:26 PM, "William Bajzek" <wi...@gmail.com> wrote:

> Hi Eric,
> 
> I'm not sure how many people are really using MySQL in this capacity. I know
> that one of the other people who has posted to the mailing list about i
> recently is my coworker, so it's possible that our recent activity is giving
> the impression of more demand than there really is. Our client is a big
> corporation who plans to use Chukwa this way and I'm working on a patch for
> them that does as you say. It is pretty small and straightforward. What would
> be a good way to contribute it back to the community?
> 
> 
> 
> On Jul 23, 2010, at 9:58 AM, Eric Yang wrote:
>> 
>> MetricDataLoader can be modified to throw IOException to the executor class
>> MetricDataLoaderPool, which can throw exception to PostProcessor Manager.
>> PostProcessorManager moves the data to a temp directory.  The retry logic
>> can be added to PostProcessorManager by counting the number of retry with
>> the error out sequence file before sending it to InError directory.  It
>> should be the better route to manages error conditions.
>> 
>> It seems like a lot of people are using MySQL part of Chukwa.  I am
>> wondering what are the use cases?  If we can have a driver on this part of
>> the project, it could be better supported without fragment the community.
> 
> 
> 
> - William Bajzek
> williambajzek@gmail.com
> 
> 
> 
> 
>