[
https://issues.apache.org/jira/browse/CHUKWA-307?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12721003#action_12721003
]
Eric Yang commented on CHUKWA-307:
----------------------------------
Chukwa records are not uniform. For example, let's say top, iostat, sar are
groupped into "system metrics". Chukwa records may contain only metrics from
iostat and sar for one record, then another record contains top. Records are
arriving at different demux which prevent this to possible. In addition,
relational database has fixed schema, it's hard to add additional columns into
database due to the fact that adding columns are time consuming process. The
real fix is to serve data from HDFS with index (CHUKWA-22), or use a column
stored system like hbase.
The goal of this jira is to focus on small deployment to reduce the complexity
of deploy chukwa without having to carry the weight of mysql.
> Refactor database to use lightweight scheme
> -------------------------------------------
>
> Key: CHUKWA-307
> URL: https://issues.apache.org/jira/browse/CHUKWA-307
> Project: Hadoop Chukwa
> Issue Type: Improvement
> Components: Data Processors, User Interface
> Reporter: Jiaqi Tan
>
> Replace MDL + MySQL with lightweight JavaDB or other portable open-source
> database.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
