[TYPO3-Solr] Index Queue Worker: error when started automatically in scheduler, no error when started manually
Andreas Brunn
brunn at archaeomedia.de
Tue Feb 28 17:53:14 CET 2012
Hi,
I just upgraded from tx_solr 1.3 to 2.0 from TER.
SOLR search is working in general after I set up the SOLR server 3.5 and
did all the necessary configuration. Pages and news get indexed when I
run the Index Queue Worker task manually and searching in the frontend
is working.
The problem is that the Index Queue Worker task in scheduler seems to
run without errors ONLY when I start it manually. Everytime when started
automatically via cron it indicates the following error and indexing
does not proceed any further:
"Ausführung fehlgeschlagen: 1311080805, Could not create a valid URL to
get frontend data while trying to index a page. Created URL:
http://www.archaeologie-online.de/index.php?id=1932&L=0"
The URL looks perfectly valid to me and when I access that URL via
browser the correct page is shown with HTTP status 200 OK.
The page id mentioned in the error message is not always the same, so it
should not be a problem with a specific page.
Could some point me to the right direction how I can fix that?
Here is some information from dev log that may give a hint to the source
of the problem:
-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
*indexQueueUid 127958
*rootPageUid 40
*type pages
*indexingConfigurationName pages
*changed 1302897750
*indexingProperties EMPTY!
*indexingPropertiesLoaded 1
*hasIndexingProperties
*recordUid 1932
constructed URL http://www.archaeologie-online.de/index.php?id=1932&L=0
scheme http
host www.archaeologie-online.de
path /
page ID 1932
indexer options EMPTY!
-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
indexingProperties and indexer options are empty. Could that cause the
problem?
If so, how would I set indexing properties and indexer options correctly
(current TS setup see below)?
As mentioned above this error occurs only when Index Queue Worker task
is started automatically by the scheduler. Other scheduler tasks don't
show any errors when run automatically.
The "Apache SOLR (solr)" template is included in website's root page
(that has "root page" enabled and a correct domain record). Custom
Typoscript is set up as follows:
-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
plugin.tx_solr {
enabled = 1
enableDebugMode = 1
useCurlHttpTransport = 0
}
plugin.tx_solr.solr {
host = localhost
port = 1050
path = /solr/core_de/
scheme = http
}
plugin.tx_solr.index {
files = 1 # enable file indexing
files.allowedTypes = doc, docx, pdf
fieldProcessingInstructions {
changed = timestampToIsoDate
created = timestampToIsoDate
endtime = timestampToIsoDate
}
}
plugin.tx_solr.index.queue {
tt_news = 1
tt_news {
fields {
abstract = short
author = author
description = short
title = title
content = SOLR_CONTENT
content {
field = bodytext
ifEmpty.cObject = TEXT
ifEmpty.cObject {
field = short
}
}
category_stringM = SOLR_RELATION
category_stringM {
localField = category
multiValue = 1
}
keywords = SOLR_MULTIVALUE
keywords {
field = keywords
}
url = TEXT
url {
typolink.parameter = {$plugin.tt_news.singlePid}
typolink.additionalParams =
&tx_ttnews[tt_news]={field:uid}&L={field:__solr_index_language}
typolink.additionalParams.insertData = 1
typolink.returnLast = url
typolink.useCacheHash = 1
}
sortAuthor_stringS = author
sortTitle_stringS = title
}
attachments {
fields = news_files
}
additionalWhereClause = type = 0
}
}
plugin.tx_solr.search {
query.allowEmptyQuery = 1
targetPage = 50
results {
resultsHighlighting = 1
siteHighlighting = 1
showDocumentScoreAnalysis = 1
markResultTypeBoundaries = 1
}
spellchecking = 1
lastSearches = 1
frequentSearches = 1
sorting = 1
sorting {
defaultOrder = asc
options {
relevance {
field = relevance
label = Relevanz
}
title {
field = sortTitle_stringS
label = Titel
}
type {
field = type
label = Typ
}
author {
field = sortAuthor_stringS
label = Autor
}
created {
field = created
label = Datum
}
}
}
faceting = 1
faceting {
# Gruppierung entfernen Link Beschriftung
removeFacetLinkText = @facetText (entfernen)
facets {
type {
label = Filtertyp
renderingInstruction = CASE
renderingInstruction {
key.field = optionValue
pages = TEXT
pages.value = Pages
pages.lang.de = Seiten
tt_news = TEXT
tt_news.value = News
tt_news.lang.de = Nachrichten
}
}
}
}
}
plugin.tx_solr.statistics = 1
plugin.tx_solr.suggest = 1
plugin.tx_solr.logging {
exceptions = 1
indexing {
indexQueueInitialization = 1
missingTypo3SearchMarkers = 1
pageIndexed = 1
queue {
pages = 1
tt_news = 1
}
}
query {
filters = 1
searchWords = 1
queryString = 1
rawPost = 1
rawGet = 1
}
}
plugin.tx_solr.templateFiles {
search = fileadmin/templates/solr/search.htm
results = fileadmin/templates/solr/results.htm
pagebrowser = fileadmin/templates/solr/pagebrowser.htm
}
plugin.tx_solr.cssFiles {
results = fileadmin/templates/solr/results.css
suggest = fileadmin/templates/solr/jquery-ui.custom.css
}
-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
Some further configuration information:
Apache/2.2.21 on Linux
Apache Tomcat/6.0.20
Apache SOLR 3.5
Java VM 1.6.0_06-b02
PHP 5.3.10
TYPO3 4.5.10
Thank you very much for taking the time to read this long post - I tried
to provide all information that seemed to be useful for tracking down
the problem.
Regards,
Andreas
--
Andreas Brunn
brunn at archaeomedia.de
More information about the TYPO3-project-solr
mailing list