[
https://issues.apache.org/jira/browse/METRON-295?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15631182#comment-15631182
]
ASF GitHub Bot commented on METRON-295:
---------------------------------------
Github user ottobackwards commented on the issue:
https://github.com/apache/incubator-metron/pull/338
You can solve this problem ( and the grok problem ) without storing the
world in ZK with tooling. An official set of scripts/web interface for
deploying scripts/topologies/grok rules to whatever location. This tooling can
change the version, checksum, crc whatever in the configurations in ZK and
trigger the reload from disk.
I don't know what the real benefit from storing the scripts or rules in ZK
are beyond notification. It seems to me to make things harder.
If there is no other reason, and it is only for triggering change maybe a
tool/flag approach would be simpler?
> Script parsing bolt
> --------------------
>
> Key: METRON-295
> URL: https://issues.apache.org/jira/browse/METRON-295
> Project: Metron
> Issue Type: New Feature
> Affects Versions: 0.2.1BETA
> Reporter: James Sirota
> Assignee: Karthik Narayanan
> Priority: Minor
> Labels: newbie, platform
> Fix For: 0.2.2BETA
>
>
> In addition to having a Grok parsing bolt we need a bolt that can execute a
> script in order to parse a telemetry. This way you can still script the
> parsing for telemetries for which Grok expressions are too complex, but still
> don't have to define a java parser
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
