컴퓨터 사용 중 메모리 부족 현상이 발생하는 경우 조치 방법

컴퓨터 자체의 사양이 증가하면서 사실상 메모리 부족이 발생하는 경우는 거의 없습니다.

그러나 산업용 프로그램의 경우 오랜시간 켜있다가 보니, 디스크의 클러스터 불량등으로 인한 문제나, 메모리 자체의 문제가 발생하면서 메모리 부족 현상이 발생하는 경우가 꽤 있습니다.

--------------------------------

이런 메모리 부족메세지와 함께 컴퓨터의 성능이 저하되는 문제가 발생하는 경우에는 메모리 자체의 문제일 수 있으므로, 메모리를 검사해 보아야 합니다.

그리고, 메모리 슬럿이나 메인보드의 슬럿에 문제가 발생하면서 메모리 부족 현상이 발생하는 경우도 있습니다.

그밖에 하드디스크의 클러스터에 문제가 있어서 연속적인 가상메모리 공간에 문제가 있는 경우에도 메모리 부족이 발생 할 수 있습니다.

이런 경우는 보통 조각모음을 해주면 문제가 해결됩니다.

2기가 바이트 이상의 메모리가 필요한 경우에는 가상메모리를 사용하면서 연속적인 클러스터 부분이 필요한 경우가 있습니다.

--------------------------------------

최대한 설정할 수 있는 만큼의 가상메모리 공간을 확보해 주는 것이 좋습니다.

특히 SSD를 사용하는 경우에는 이런 문제가 하드디스크 보다 빈번하게 발생하는 것이 아닌가 생각됩니다.

랩뷰 메모리 부족 현상에 대해서

https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z000000P9mtSAC

Issue Details

When I am developing a large application in LabVIEW, I sometimes receive the following error message, even though I have a large amount of available RAM.

LabVIEW: Memory is full.
The top-level VI was stopped on the block diagram.
Refer to the VI Memory Usage topic in the LabVIEW Help for suggestions on monitoring and improving VI memory usage.

Solution

You typically receive this message when you are developing an application with large data sets of arrays, clusters, or waveforms. Because LabVIEW stores each data set in a contiguous block of memory, the block of memory available on your computer might not be large enough to store a large data set.

Use the suggestions in the following sections to prevent this message from appearing:

Reducing the Memory Arrays, Clusters, and Waveforms Use

• If you have an array with a large data set, wire a smaller data type to the array, if possible. For example, if the array is a double-precision, floating point number [DBL] and you are storing integers in the 0 to 100 range, you can change the integer representation to Unsigned 8 bit integers [U8]. A U8 data type uses an eighth of the memory that a double-precision floating point number uses.
• In LabVIEW 8.5 and later, use the In Place Element Structure to avoid making copies of data when using an array, cluster, or waveform.
• If you are graphing data from a large array, use max-min decimation to avoid graphing every data point. Refer to LabVIEW Help: Memory Management for Large Data Sets for more information about max-min decimation.
• Analyze the large data set to determine where memory is full. Stream or write the data to disk and use a high-performance analysis and reporting tool, such as TDMS files, and NI DIAdem software.

Reducing Memory Usage in Applications

• To limit the number of times LabVIEW copies data, reduce the number of global and local variables in an application. Instead of using variables, you can create loops using a producer/consumer design pattern, or you can use the New Data Value Reference function to create a reference to the data.
• Separate sections of code into subVIs to make memory available when the subVIs are not executing.
• Use the Call By Reference Node to call large subVIs dynamically. A subVI is not in memory until a VI calls it dynamically.
• Use the Profile Performance and Memory Window to acquire and display data about execution and data time. How Can I Monitor Execution Time and Memory Usage in My LabVIEW VIs?
• Make improvements to the code to optimize memory usage. Refer to How Can I Optimize the Memory Use in My LabVIEW VI? for more information about memory optimization.
• Use the NI LabVIEW Desktop Execution Trace Toolkit to identify problems that could negatively impact performance such as memory leaks and reference leaks; collect low-level information such as call chain, thread ID, and location in virtual memory of trace; as well as profile VIs, executables, and shared libraries.

Increasing Memory in LabVIEW

• Increase the amount of virtual memory on the system. Refer to How Can Virtual Memory Help With Memory Issues in LabVIEW? for more information about virtual memory.
• The amount of memory LabVIEW can utilize depends on the version and operating system used. Refer to How Much Memory Can LabVIEW Use? for more information about increasing memory.

 

Additional Information

This message is most common if LabVIEW cannot allocate a block of memory large enough for an array. For example, an array of double-precision, floating point numbers with 2,000 by 2,000 elements uses approximately 32 MB. Even if 32 MB of memory is available, you may not have a block of memory large enough for the array.
In the specific example seen above, this error is generated when an array containing more than 2^31 elements is used to index a For Loop. Arrays can contain up to 2^32 elements, but a For Loop can only index up to 2^31 values. This is because the N terminal uses the I32 datatype rather than the U32. The most significant bit of the I32 datatype is used to specify whether the value is positive or negative.