hadoop-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Konstantin Shvachko <...@yahoo-inc.com>
Subject Re: HDFS name node HA roadmap
Date Fri, 19 Jun 2009 17:44:22 GMT
Hi Jie Qiu,

It would be interesting to see what your approach to HA is.
Indeed, I had some ideas on how to implement it.
And I would be happy to talk to you about it.

Which jira do you refer to?

Thanks,
--Konstantin

Jie Qiu wrote:
> 
> Hi, Dhruba,
> 
> Yes, we plan to open a JIRA to let community to review our 
> implementation and give comments. Recently, we are going with IBM 
> internal process before submitting to the community.
> 
> I also want to discuss with shv whether we need to another separate JIRA 
> or work on the same integrated JIRA.
> 
> Do you any suggestion on this?
> 
> ==================================================
> True insight comes from within!
> 
> Qiu Jie
> 
> Autonomic Middleware and Service Delivery
> IBM China Research Lab.
> Tel: 8610-58748086  Tieline: 905-8086
> FAX: 8610-58748330          
> Add.: Building 19 Zhongguancun Software Park, 8 Dongbeiwang WestRoad, 
> Haidian District, Beijing,P.R.C.100094
> Email: qiujie@cn.ibm.com
> ==================================================
> 
> 
> *Dhruba Borthakur <dhruba@gmail.com>*
> 
> 2009-06-20 01:19
> Please respond to
> general@hadoop.apache.org
> 
> 
> 	
> To
> 	general@hadoop.apache.org
> cc
> 	shv@yahoo-inc.com, andrew.wharton@nokia.com
> Subject
> 	Re: HDFS name node HA roadmap
> 
> 
> 	
> 
> 
> 
> 
> 
> Hi Hie,
> 
> If you are working on this one, it woudl be nice if you can open a JIRA and
> let the community comment/review your design. This process usually takes a
> while and it is good to get a an early start on it.
> 
> thanks,
> dhruba
> 
> 
> On Fri, Jun 19, 2009 at 9:23 AM, Jie Qiu <qiujie@cn.ibm.com> wrote:
> 
>  > Hi Andrew,
>  >
>  > Yes, our IBM is implementing HA for Hadoop which will be open sourced
>  > soon. I think shv from Yahoo also post JIRA to describe this issue.
>  >
>  > However, our approach may be different with shv. Let me discuss with shv
>  > about how to combine our thoughts and implementation together.
>  >
>  > ==================================================
>  > True insight comes from within!
>  >
>  > Qiu Jie
>  >
>  > Autonomic Middleware and Service Delivery
>  > IBM China Research Lab.
>  > Tel: 8610-58748086  Tieline: 905-8086
>  > FAX: 8610-58748330
>  > Add.: Building 19 Zhongguancun Software Park, 8 Dongbeiwang WestRoad,
>  > Haidian District, Beijing,P.R.C.100094
>  > Email: qiujie@cn.ibm.com
>  > ==================================================
>  >
>  >
>  >
>  > Andrew Wharton <andrew.wharton@nokia.com>
>  > 2009-06-20 00:10
>  > Please respond to
>  > general@hadoop.apache.org
>  >
>  >
>  > To
>  > general@hadoop.apache.org
>  > cc
>  >
>  > Subject
>  > HDFS name node HA roadmap
>  >
>  >
>  >
>  >
>  >
>  >
>  > Hello,
>  >
>  > My understanding is that the HDFS name node single point of failure will
>  > be removed in Hadoop 0.21. Is that still the case? Also, what is the
>  > expected delivery date of that release?
>  >
>  > Thanks.
>  >
>  > -- Andrew
>  >
>  >
>  >
> 

Mime
View raw message