Has this ever worked? IOW, is this something that's changed or has just never worked?
The obvious first step is to start Solr with more than 1G of memory. Solr _likes_ memory and a 1G heap is quite small. But you say: "Increasing the heap size further doesnt start SOLR instance itself.". How much RAM do you have on your machine? What other programs are running? You should be able to increase the heap and start Solr if you have the RAM on your machine so I'd figure out what's behind that issue first. I regularly start Solr with 16 or 32G of memory on my local machines, I know of installations running Solr with 60G heaps so this points to something really odd about your environment. When you "increase it further", exactly _how_ does Solr fail to start? What appears in the Solr logs? etc. Really, troubleshoot that issue first I'd recommend. If DIH still needs a ridiculous amount of memory, it's usually the JDBC driver trying to read all the rows into memory at once and you'll have to explore the jdbc driver settings in detail. Best, Erick On Tue, Sep 18, 2018 at 11:16 PM Tanya Bompi <tanya.bo...@gmail.com> wrote: > > Hi, > I am using the Microsoft Jdbc driver 6.4 version in Solr 7.4.0 . I have > tried removing the selectMethod=Cursor and still it runs out of heap space. > Do we have anyone who has faced similar issue. > > Thanks > Tanya > > > On Tue, Sep 18, 2018 at 6:38 PM Shawn Heisey <apa...@elyograg.org> wrote: > > > On 9/18/2018 4:48 PM, Tanya Bompi wrote: > > > I have the SOLR 7.0 setup with the DataImportHandler connecting to the > > > sql server db. I keep getting OutOfMemory: Java Heap Space when doing a > > > full import. The size of the records is around 3 million so not very > > huge. > > > I tried the following steps and nothing helped thus far. > > > > See this wiki page: > > > > https://wiki.apache.org/solr/DataImportHandlerFaq > > > > You already have the suggested fix -- setting responseBuffering to > > adaptive. You might try upgrading the driver. If that doesn't work, > > you're probably going to need to talk to Microsoft about what you need > > to do differently on the JDBC url. > > > > I did find this page: > > > > > > https://docs.microsoft.com/en-us/sql/connect/jdbc/using-adaptive-buffering?view=sql-server-2017 > > > > This says that when using adaptive buffering, you should avoid using > > selectMethod=cursor. So you should try removing that parameter. > > > > Thanks, > > Shawn > > > >
