"Invalid UUID string" in Solr search core after DSpace 6 upgrade

[Alan Orth]

Dear list,

I'm testing DSpace version 5.8 to 6.3 upgrade. Immediately after performing the database migrations successfully and starting Tomcat I initiated a `dspace index-discovery -b`. I thought something was strange because I didn't see any indexing progress in dspace.log (despite high CPU usage), and then when I browsed the home page I saw many errors such as this in dspace.log:

2020-02-03 10:27:14,485 ERROR org.dspace.browse.ItemCountDAOSolr @ caught exception: 
org.dspace.discovery.SearchServiceException: Invalid UUID string: 1
2020-02-03 13:20:20,475 ERROR org.dspace.app.xmlui.aspect.discovery.AbstractRecentSubmissionTransformer @ Caught SearchServiceException while retrieving recent submission for: home page
org.dspace.discovery.SearchServiceException: Invalid UUID string: 111210

I checked Solr's search core and saw that the search.resourceid fields were indeed integers instead of UUIDs. On a guess I stopped the indexing, deleted the search cores¹, touched an empty file in dspace/solr/search/conf/reindex.flag², and restarted Tomcat. After Tomcat started back up the re-indexing started and correctly indexed search.resourceid as UUIDs.

After the initial reindex.flag indexing completes, subsequent manual indexing runs with `dspace index-discovery -b` work correctly. This only seems to happen the first time after upgrade—I tried deleting all search documents and starting a `dspace index-discovery -b` and it behaved correctly.

Has anyone seen this? Is it worth filing a bug? Are there perhaps some fixes in the as-of-yet-unreleased 6.4 that would address this?

Regards,

¹ curl -s 'http://localhost:8080/solr/search/update?stream.body=<delete><query>*:*</query></delete>&commit=true'

² https://wiki.lyrasis.org/display/DSDOC6x/Upgrading+DSpace

--

Alan Orth
alan...@gmail.com
https://picturingjordan.com
https://englishbulgaria.net
https://mjanja.ch
"In heaven all the interesting people are missing." ―Friedrich Nietzsche

[Mark H. Wood]

According to https://wiki.lyrasis.org/display/DSDOC6x/Upgrading+DSpace
the 'search' core should have been reindexed automatically, unless you
disabled that (as discussed further down that page) and then did not
follow through with a manual reindex.

You should also be aware of a usage statistics issue which will be
addressed in 6.4:

https://wiki.lyrasis.org/display/DSDOC6x/SOLR+Statistics+Maintenance#SOLRStatisticsMaintenance-UpgradeLegacyDSpaceObjectIdentifiers(pre-6xstatistics)toDSpace6xUUIDIdentifiers
https://jira.lyrasis.org/browse/DS-4110

--
Mark H. Wood
Lead Technology Analyst

University Library
Indiana University - Purdue University Indianapolis
755 W. Michigan Street
Indianapolis, IN 46202
317-274-0749
www.ulib.iupui.edu

[Alan Orth]

Dear Mark,

Is there a difference in the automatic re-indexing that is triggered by the existence of the `dspace/solr/search/conf/reindex.flag` and a manually triggered `dspace index-discovery -b`? After my database migration was completed I started Tomcat and then immediately triggered a manual re-indexing. I lost a few hours investigating the cause of this, only to stumble on the fact that the "automatic" re-indexing apparently does something differently—the effects of which are only visible if the "automatic" re-indexing actually completes without interruption! Very strange...

Regards,

--
All messages to this mailing list should adhere to the DuraSpace Code of Conduct: https://duraspace.org/about/policies/code-of-conduct/
---
You received this message because you are subscribed to the Google Groups "DSpace Technical Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dspace-tech...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dspace-tech/20200204143736.GB14849%40IUPUI.Edu.

[Mark H. Wood]

On Tue, Feb 04, 2020 at 07:25:31PM +0200, Alan Orth wrote:
> Is there a difference in the automatic re-indexing that is triggered by the
> existence of the `dspace/solr/search/conf/reindex.flag` and a manually
> triggered `dspace index-discovery -b`? After my database migration was
> completed I started Tomcat and then immediately triggered a manual
> re-indexing. I lost a few hours investigating the cause of this, only to
> stumble on the fact that the "automatic" re-indexing apparently does
> something differently—the effects of which are only visible if the
> "automatic" re-indexing actually completes without interruption! Very
> strange...

I think the relevant code is in
org.dspace.storage.rdbms.DatabaseUtils.ReindexerThread

It appears to be doing the equivalent of 'dspace index-discovery -c;
dspace index-discovery -b; dspace index-discovery -s'.

(Aside: IndexClient's option processing could be better. For
example, 'index-discovery -c -b -s' would run, ignoring -b and -s. It
should complain that -c and -b are mutually exclusive.
'index-discovery -b -s' would run, ignoring -s, but there's no reason
not to obey both (though they should be sequenced: first -b and then
-s).)