[
https://issues.apache.org/jira/browse/SOLR-7038?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Anshum Gupta updated SOLR-7038:
-------------------------------
Attachment: SOLR-7038.patch
I've manually tested this patch using the scripts. Currently running the tests.
This validates the config set (presence of one) and returns and exception if
there's no config with the specified/derived name.
> If no configset exists, CREATE leads to a 500 error with never-ending logging
> and 100% CPU usage
> ------------------------------------------------------------------------------------------------
>
> Key: SOLR-7038
> URL: https://issues.apache.org/jira/browse/SOLR-7038
> Project: Solr
> Issue Type: Bug
> Affects Versions: 5.0
> Reporter: Anshum Gupta
> Assignee: Anshum Gupta
> Priority: Blocker
> Fix For: 5.0, Trunk
>
> Attachments: SOLR-7038.patch
>
>
> Here's what I did:
> {code}
> > bin/solr start -e cloud -noprompt
> > curl
> > http://localhost:8983/solr/admin/collections?action=CREATE&name=thisshouldfail&numShards=1&configset=thisisaninvalidconfigset&wt=json
> {code}
> The above led to a new collection named thisshouldfail, with the config-set
> as gettingstarted. This call should have failed as there was no configset by
> that name. Instead, it picked up the only config set it found and used it.
> There's more to this. I'm not sure how related this is but looks like it to
> me.
> {code}
> > bin/solr start -c
> > curl
> > http://localhost:8983/solr/admin/collections?action=CREATE&name=thisshouldfail&numShards=1&configset=thisisaninvalidconfigset&wt=json
> {code}
> This led to a 900M (and growing) log file in addition to 100% CPU until I
> killed Solr.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
