You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Erick Erickson (JIRA)" <ji...@apache.org> on 2013/10/06 16:48:41 UTC

[jira] [Comment Edited] (SOLR-5253) Move _version_ (and other _*_ fields) to right above the id field in the example schema.xml

    [ https://issues.apache.org/jira/browse/SOLR-5253?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13787625#comment-13787625 ] 

Erick Erickson edited comment on SOLR-5253 at 10/6/13 2:47 PM:
---------------------------------------------------------------

Reality is that if you try to start without _version_ in your schema Solr doesn't start. Stack traces all over the place and a non-functioning installation, can't query, can't get to localhost:8983:/solr, etc. At least that's what happened when I just tried it on trunk. Top-level error below. Or I screwed up again when I tested before committing this, that's been known to happen.

So the current comments _do_ reflect "reality". Whether it should behave this way is a different question, I was a bit surprised by this behavior as well. Perhaps another JIRA about making _version_ not required in non-cloud mode is in order?

Hmmm, or I need to disable update log..... Let me try that.

As for "id", as far as I know it's not required. Personally I'd like to _make_ it required since, as you say, so much breaks if you don't have it. But that's another JIRA too.

{msg=SolrCore 'collection1' is not available due to init failure: Unable to use updateLog: _version_ field must exist in schema, using indexed="true" stored="true" and multiValued="false" (_version_ does not exist),trace=org.apache.solr.common.SolrException: SolrCore 'collection1' is not available due to init failure: Unable to use updateLog: _version_ field must exist in schema, using indexed="true" stored="true" and multiValued="false" (_version_ does not exist)
	at org.apache.solr.core.CoreContainer.getCore(CoreContainer.java:785)




was (Author: erickerickson):
Reality is that if you try to start without _version_ in your schema Solr doesn't start. Stack traces all over the place and a non-functioning installation, can't query, can't get to localhost:8983:/solr, etc. At least that's what happened when I just tried it on trunk. Top-level error below. Or I screwed up again when I tested before committing this, that's been known to happen.

So the current comments _do_ reflect "reality". Whether it should behave this way is a different question, I was a bit surprised by this behavior as well. Perhaps another JIRA about making _version_ not required in non-cloud mode is in order?

As for "id", as far as I know it's not required. Personally I'd like to _make_ it required since, as you say, so much breaks if you don't have it. But that's another JIRA too.

{msg=SolrCore 'collection1' is not available due to init failure: Unable to use updateLog: _version_ field must exist in schema, using indexed="true" stored="true" and multiValued="false" (_version_ does not exist),trace=org.apache.solr.common.SolrException: SolrCore 'collection1' is not available due to init failure: Unable to use updateLog: _version_ field must exist in schema, using indexed="true" stored="true" and multiValued="false" (_version_ does not exist)
	at org.apache.solr.core.CoreContainer.getCore(CoreContainer.java:785)



> Move _version_ (and other _*_ fields) to right above the id field in the example schema.xml
> -------------------------------------------------------------------------------------------
>
>                 Key: SOLR-5253
>                 URL: https://issues.apache.org/jira/browse/SOLR-5253
>             Project: Solr
>          Issue Type: Improvement
>    Affects Versions: 4.5, 5.0
>            Reporter: Erick Erickson
>            Assignee: Erick Erickson
>            Priority: Minor
>         Attachments: SOLR-5253.patch, SOLR-5253.patch
>
>
> Minor, but it bugs me that it's so easy to try to remove all extraneous fields from schema.xml and shoot yourself in the foot. Now and forever more we should place all the "special" fields right at the top of the example schema. Trivial to do.
> True, we say a nice note "_*_ fields are internal and required", but still.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org