Yeah fine..I will do that...Before the merge Factor was 10 itself ...After finding this error I just set its value higher assuming if that could be error anyway... Will re change it..
The ramBufferSize is 256MB... Do I need to change this value to higher? On Wed, Aug 18, 2010 at 12:27 AM, Jay Hill <jayallenh...@gmail.com> wrote: > A merge factor of 100 is very high and out of the norm. Try starting with a > value of 10. I've never seen a running system with a value anywhere near > this high. > > Also, what is your setting for ramBufferSizeMB? > > -Jay > > On Tue, Aug 17, 2010 at 10:46 AM, rajini maski <rajinima...@gmail.com > >wrote: > > > yeah sorry I forgot to mention others... > > > > <mergeFactor>100</mergeFactor> > > <maxBufferedDocs>1000</maxBufferedDocs> > > <maxMergeDocs>100000</maxMergeDocs> > > <maxFieldLength>10000</maxFieldLength> > > > > above are the values.... > > > > Is this because of values here...initially I had mergeFactor parameter > -10 > > and maxMergedocs-10000....With the same error i changed them to above > > values..Yet I got that error after index was about 2lacs docs... > > > > On Tue, Aug 17, 2010 at 11:04 PM, Erick Erickson < > erickerick...@gmail.com > > >wrote: > > > > > There are more merge paramaters, what values do you have for these: > > > > > > <mergeFactor>10</mergeFactor> > > > <maxBufferedDocs>1000</maxBufferedDocs> > > > <maxMergeDocs>2147483647</maxMergeDocs> > > > <maxFieldLength>10000</maxFieldLength> > > > > > > See: http://wiki.apache.org/solr/SolrConfigXml > > > > > > Hope that formatting comes through the various mail programs OK.... > > > > > > Also, what else happens while you're indexing? Do you search > > > while indexing? How often do you commit your changes? > > > > > > > > > > > > On Tue, Aug 17, 2010 at 1:18 PM, rajini maski <rajinima...@gmail.com> > > > wrote: > > > > > > > <mergefactor>100 </mergefactor> > > > > JVM Initial memory pool -256MB > > > > Maximum memory pool -1024MB > > > > > > > > <add> > > > > <doc> > > > > <field>long:ID</field> > > > > <field>str:Body</field> > > > > .... > > > > ....12 fields > > > > </filed> > > > > </doc> > > > > </add> > > > > I have a solr instance in solr folder (D:/Solr) free space in disc is > > > > 24.3GB > > > > .. How will I get to know what portion of memory is solr using ? > > > > > > > > > > > > > > > > On Tue, Aug 17, 2010 at 10:11 PM, Erick Erickson < > > > erickerick...@gmail.com > > > > >wrote: > > > > > > > > > You shouldn't be getting this error at all unless you're doing > > > something > > > > > out of the ordinary. So, it'd help if you told us: > > > > > > > > > > >What parameters you have set for merging > > > > > >What parameters you have set for the JVM > > > > > >What kind of documents are you indexing? > > > > > > > > > > The memory you have is irrelevant if you only allocate a small > > > > > portion of it for the running process... > > > > > > > > > > Best > > > > > Erick > > > > > > > > > > On Tue, Aug 17, 2010 at 7:35 AM, rajini maski < > rajinima...@gmail.com > > > > > > > > wrote: > > > > > > > > > > > I am getting it while indexing data to solr not while querying... > > > > > > Though I have enough memory space upto 40GB and I my indexing > data > > is > > > > > just > > > > > > 5-6 GB yet that particular error is seldom observed... (SEVERE > > ERROR > > > : > > > > > JAVA > > > > > > HEAP SPACE , OUT OF MEMORY ERROR ) > > > > > > I could see one lock file generated in the data/index path just > > after > > > > > this > > > > > > error. > > > > > > > > > > > > > > > > > > > > > > > > On Tue, Aug 17, 2010 at 4:49 PM, Peter Karich <peat...@yahoo.de> > > > > wrote: > > > > > > > > > > > > > > > > > > > > > Is there a way to verify that I have added correctlly? > > > > > > > > > > > > > > > > > > > > > > on linux you can do > > > > > > > ps -elf | grep Boot > > > > > > > and see if the java command has the parameters added. > > > > > > > > > > > > > > @all: why and when do you get those OOMs? while querying? which > > > > queries > > > > > > > in detail? > > > > > > > > > > > > > > Regards, > > > > > > > Peter. > > > > > > > > > > > > > > > > > > > > > > > > > > > >
