JVM Error possibly due to memory error

3.11K viewsGeneral Discussion
0

Hi:

I just installed Quantrix and it has problem starting up. I get an error message that says “The JVM could not be started. The main method may have thrown an exception.”

Prior to failed start up, I was able to start up the application until it crashedix crashed while opening a model that is just over 4MB in size. I guess it ran out of memory. The model uses more than 400MB of memory based on the Task Manager. So, I increased the memory setting in the application to 1.5 gig (I have 5 gig of memory in my machine.) and it took the entry. Now, Quantrix refused to start up and I can’t change the memory to something lower because I can’t start the application.

Does this require a new install? Thanks.

Henry

0

Henry,

I looks like you have exceeded a limit. You can reset this outside Quantrix by editing the file Quantrix Modeler.exe.vmoptions in the Quantrix program directory (usually C:Program FilesQuantrix Modeler). Edit the line that reads “-Xmx1500M” to be a lower number, save the file and then try starting Quantrix.

Let us know if you need further assistance.

Chris

0

Hi Chris:

Thanks for the quick response.

I had to bring down the memory to 1.1 Gig to get the Quantrix started.

I still very troubled by how the memory is used for Quantrix because it severely limits the type of analysis or data that it can handle. The Task Manager was showing that I was using 700MB of memory while opening 4MB model. Once the model was opened, the memory usage went down to 36Mb. That is a huge difference!

I don’t know why there is such high memory requirement with certain uses, like opening the model, the datalink, and copy/paste. But, is this a bug that can be fixed? (It seems like the model can be opend in subparts to avoid running out of memory….)

When I had problems with the size of ODBC Datalink, I had thought doing multiple imports to avoide the memory issue during the import. (My understanding is that the memory limit was more due to the loading process, not the size of the data.) But, it looks like multiple batch load may not work because even if all the data can be loaded, one may not be able to open the model.

The other thing that I noticed is that the memory used by Quantrix only drops when the application is minimized. Othewise, it is pegged at 1 gig after the import. If the application is minized then brought back up the memory in use goes down to 40MB.

Can you please elaborate whether the memory issue is something that can be addressed with a patch? Is the memory usage related to using compressed XML file format? (I am guessing because it can take several minutes to save a 7MB file using a Xeon 3.6 Ghz and 5 Gig memory system……)

Best regards,

Henry

0

Henry,

All of these points are relevant and ones that we have heard before with other users. We are currently developing plans to address these issues with our next major release of Quantrix. So I don’t have anything immediate to assist you with your concerns, but please be advised we are keeping these in mind for our next major release.

-Mike

0

Mike:

Thank you for your prompt response as always.

I am glad that you are thinking of “scaled-up” Quantrix in the next release. I am very excited about what Quantrix offers at this point except for scale-up issues that I am running into. Though I am new to Quantrix, I think I have the appreciation for what you guys are trying to do with Quantrix and see much promise in various applications. I actually see Quantrix to be very applicable to large-scale or data-intensive models where you need something more than what Exel or ACCESS can offer. So, I want to strongly encourage you to consider scaling up Quantrix. I will be looking for Quantrix’s roadmap for the next release. Thank you.

Best regards,

Henry

Latest Questions