[
https://issues.apache.org/jira/browse/KNOX-502?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15120424#comment-15120424
]
Kevin Minder commented on KNOX-502:
-----------------------------------
After further investigation, the above audit log sample isn't really wrong it
is just misleading. I'm making a fix to add auditing to the servlet in Knox
that does the forwarding for the default topology. The results in an audit log
that looks like this so that it is clear that a forward occurred hopefully
avoiding any confusion.
{code}
16/01/27 18:24:18
||429ff847-fc69-4029-a7bf-03d6762634b1|audit|||||forward|uri|/gatewayy/sandbox/webhdfs/v1/|unavailable|Forward
method: GET to default context: /gateway/sandbox
16/01/27 18:24:18
||429ff847-fc69-4029-a7bf-03d6762634b1|audit|||||access|uri|/gateway/sandbox/gatewayy/sandbox/webhdfs/v1/?op=GETHOMEDIRECTORY|unavailable|Request
method: GET
16/01/27 18:24:18
||429ff847-fc69-4029-a7bf-03d6762634b1|audit|||||access|uri|/gateway/sandbox/gatewayy/sandbox/webhdfs/v1/?op=GETHOMEDIRECTORY|success|Response
status: 404
16/01/27 18:24:18
||429ff847-fc69-4029-a7bf-03d6762634b1|audit|||||forward|uri|/gatewayy/sandbox/webhdfs/v1/|success|Response
status: 404
{code}
> Invalid requests (404s) should be logged and audited
> ----------------------------------------------------
>
> Key: KNOX-502
> URL: https://issues.apache.org/jira/browse/KNOX-502
> Project: Apache Knox
> Issue Type: Improvement
> Components: Server
> Affects Versions: 0.5.0
> Reporter: Kevin Minder
> Assignee: Kevin Minder
> Fix For: 0.8.0
>
>
> Currently there is no Knox level diagnostic or audit information produced for
> 404s. This has been making it difficult for users to self diagnose
> mismatches between their perception of a cluster name and a topology name.
> We should provide better diagnostics here to help the users figure out what
> the problem is.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
