2

I have a big synonym dictionary and I created a file correctly formated to use it with solr. This file has a size of 4mb. But when I want to start solr using Tomcat, solr is blocked in start.

What is the problem ? The file is too big ?

Thanks.

bobosh
  • 425
  • 5
  • 21
  • 1
    What is the error message/stacktrace ? Can u start it without the file? – KaKa Jul 20 '11 at 00:55
  • I can start without the file and I can start if I remove some text of this file. There is no error, the console block at "INFO: unique key field :id". I feel that Solr want to start but it can't. I already increased the memory size of tomcat (Xmx1024m) but it change nothing. – bobosh Jul 20 '11 at 07:20

2 Answers2

4

we created a new synonymsfilter to address this problem, the old synonymsfilter was quite bad with such large synonyms files:

https://issues.apache.org/jira/browse/LUCENE-3233

Unless something changes, should make it into 3.4

Robert Muir
  • 3,185
  • 21
  • 17
0

I think it has nothing to do with tomcat memory, try increasing solr memory. Found this link which mentions that synonym.txt will take lot of heap space, so mostly its due to solr running out of memory

KaKa
  • 1,543
  • 12
  • 18
  • How can I increase solr memory ? In solrconifg.xml ? – bobosh Jul 20 '11 at 08:07
  • While starting solr add an option -Xmx4096 , keep checking the increase in memory to confirm whether its running out of memory while startup – KaKa Jul 20 '11 at 08:13
  • I start Solr through Tomcat... So it's why i increase memory size of Tomcat – bobosh Jul 20 '11 at 08:19
  • solr,tomcat will run as separate processes, check how much memory each one is getting, never ran solr through tomcat, hope this helps – KaKa Jul 20 '11 at 08:52