You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@chukwa.apache.org by Jerome Boulon <jb...@netflix.com> on 2010/02/24 18:41:40 UTC
Survey - Chukwa usage
Hi,
I would like to get some stats on how Chukwa is used?
Thanks in advance,
/Jerome.
* cluster status
> * personal
> * business experiment
> * business dev
> * business test
> * business prod
* cluster size
> * < 10 nodes
> * < 100 nodes
> * < 200 nodes
> * < 500 nodes
> * > 1000 nodes
* Chukwa usage
> * monitoring
> * analytics
* Are you using Chukwa to collect:
> * Log files
> * Metrics information
* Cluster running
> * in your own data center
> * in the cloud (ex: amazon ec2)
* which Chukwa components are you using?
> * UI (HICC)
> * Log4j Appender
> * Agent adaptors (list)
> * Archiver
> * Demuxer
* Data analytics on ChukwaRecords
> * pig
> * M/R
* Would you be interested in an agent less solution?
* Would you like to be able to use Hive instead of/in addition to Pig?
* What from your point of view is missing in Chukwa, today?
Re: Survey - Chukwa usage
Posted by James Seigel <ja...@tynt.com>.
On 2010-02-24, at 10:41 AM, Jerome Boulon wrote:
cluster status
business – prod
cluster size
< 100 nodes
Chukwa usage
analytics
Are you using Chukwa to collect:
Log files
Cluster running
in your own data center
which Chukwa components are you using?
Agent adaptors (list)(DirTailingAdaptor)
Archiver
Demuxer
Data analytics on ChukwaRecords
pig
Would you be interested in an agent less solution?
no
Would you like to be able to use Hive instead of/in addition to Pig?
in addition
What from your point of view is missing in Chukwa, today?
Better docs on data flow. Better plugability. Triggers
Re: Survey - Chukwa usage
Posted by Guillermo Pérez <bi...@tuenti.com>.
cluster status: business – dev
cluster size: < 10 nodes (for now)
Chukwa usage: analytics
Are you using Chukwa to collect: Custom application events (more like a log)
Cluster running: in your own data center
which Chukwa components are you using? UI (HICC), Archiver, Demuxer, BackFill
Data analytics on ChukwaRecords: pig
Would you be interested in an agent less solution? Yes. Integration
with syslog-ng
Would you like to be able to use Hive instead of/in addition to Pig? Yes
What from your point of view is missing in Chukwa, today?
- Better reports of errors during demux.
- More integrated and easy to modify framework for generating reports
and push them to mysql. HICC does this, but is hard adapt to your
needs.
--
Guille -ℬḭṩḩø- <bi...@tuenti.com>
:wq
Re: Survey - Chukwa usage
Posted by Eric Yang <ey...@yahoo-inc.com>.
Cluster Status business dev
Cluster Size 10 nodes
Chukwa Usage: monitoring + analytics
Cluster: yahoo data center
All components
Data analytics: pig+MR
Agent less solution: no preference
Hive: no preference
What is missing:
- Missing good storage format.
- Missing good post demux report generator.
- Missing good data retention process.
- Missing full body index for Archive for search, i.e. find log entries.
- The list goes on and on...
It will take time to build all those features.
Regards,
Eric
On 2/24/10 9:41 AM, "Jerome Boulon" <jb...@netflix.com> wrote:
> Hi,
> I would like to get some stats on how Chukwa is used?
> Thanks in advance,
> /Jerome.
>
> * cluster status
>> * personal
>> * business experiment
>> * business dev
>> * business test
>> * business prod
>
> * cluster size
>> * < 10 nodes
>> * < 100 nodes
>> * < 200 nodes
>> * < 500 nodes
>> * > 1000 nodes
>
> * Chukwa usage
>> * monitoring
>> * analytics
>
> * Are you using Chukwa to collect:
>> * Log files
>> * Metrics information
>
> * Cluster running
>> * in your own data center
>> * in the cloud (ex: amazon ec2)
>
> * which Chukwa components are you using?
>> * UI (HICC)
>> * Log4j Appender
>> * Agent adaptors (list)
>> * Archiver
>> * Demuxer
>
> * Data analytics on ChukwaRecords
>> * pig
>> * M/R
>
> * Would you be interested in an agent less solution?
>
> * Would you like to be able to use Hive instead of/in addition to Pig?
>
> * What from your point of view is missing in Chukwa, today?
>
>