Stack Overflow
Stack Overflow

Questions tagged [numa]

NUMA stands for Non Uniform Memory Access. It is a general linux term indicating that the hardware has multiple memory nodes, and that not all processing units have equal access to all memory.

NUMA stands for Non Uniform Memory Access. It is a general linux term indicating that the hardware has multiple memory nodes, and that not all processing units have equal access to all memory.

As processors become faster and faster, proximity to memory increases in importance for overall computing performance. NUMA systems address this problem by building closer connections between specific computing resources and memory.

307 questions
0
votes
1 answer

MongoDB and NUMA

MongoDB write performance is really slow on this machine with NUMA. I went through this https://docs.mongodb.com/manual/administration/production-notes/ . I disabled zone reclaim and started mongodb with numactl --interleave=all command as…
MongoDBDeveloper
  • 41
  • 6
0
votes
1 answer

What are some computers that support NUMA?

What are some computers that support NUMA? Also, how many cores are required? I have tried searching in Google and Bing but couldn't find any answers.
pgp
  • 83
  • 6
0
votes
1 answer

How does the memory behind statically allocated huge pages get distributed across NUMA nodes?

In my /etc/default/grub file I have explicitly set aside N huge pages with "hugepages=N". If I'm running on a box with 2 NUMA nodes, does N/2 huge pages get set aside for each node, or do they all go to node 0, or ....? Also is there a way on the…
Joseph Garvin
  • 20,727
  • 18
  • 94
  • 165
0
votes
1 answer

Static array of structs with pointers

We have an array of structs like this one: struct allocation { size_t alloc_size_; char* alloc_memory_; }; static struct allocation allocations[] = {{1024, NULL},{2048, NULL},}; later on in main() it's members alloc_memory_ are initialized…
Dimon
  • 436
  • 5
  • 15
0
votes
0 answers

Cross compile static libnuma for arm

I'm trying to cross compile libnuma for arm arch. In process of compiling i am facing below issue while generating libnuma.so file. arm-linux-gnueabi-gcc -static -Wl,-soname=libnuma.so.1 -Wl,--version-script,versions.ldscript -Wl,-init,numa_init -o…
msk
  • 135
  • 2
  • 13
0
votes
1 answer

How to compile Numa library?

How should I compile the numa source code(ftp://oss.sgi.com/www/projects/libnuma/download/numactl-2.0.10.tar.gz) to get the libnuma.so, both for 32bit and 64bit?
Shaohua
  • 21
  • 1
0
votes
1 answer

Network Adapter NUMA

Is there any way in a windows system, to (a) programmatically find the NUMA node of a network adapter without writing a driver or (b) read and write the Receive Side Scaling option for a network adapter without writing a driver? The best solution…
bazz-dee
  • 687
  • 5
  • 23
0
votes
0 answers

Which numa will a dll or exe code page load into

We run multiple copies of our application at the same time on a multi core server with 2 NUMA nodes. We already set the affinity of the app to cores on either NUMA zero or one to increase memory access performance. However it occurred to me that…
johnnycrash
  • 5,184
  • 5
  • 34
  • 58
0
votes
1 answer

Mapping of memory addresses to physical modules in Windows XP

I plan to run 32-bit Windows XP on a workstation with dual processors, based on Intel's Nehalem microarchitecture, and triple channel RAM. Even though XP is limited to 4 GB of RAM, my understanding is that it will function with more than 4 GB…
Josef Grahn
  • 1,585
  • 9
  • 12
0
votes
1 answer

Memory not seen in /proc/meminfo on dual Xeon setup

I have a dual Xeon CPU setup (motherboard Asus Z8NA-D6) with 24 GB of memory (12 GB per CPU). I can't see 24GB of memory in /proc/meminfo but I can see it in /sys/devices/system/node/node0/meminfo: $ cat /proc/meminfo MemTotal: 13810952…
Lienj
  • 1
0
votes
1 answer

Page cache pile up and force memory allocation on other nodes

I am currently working on a NUMA machine. I came across an issue that when I perform disk I/O, the page cached started to pile up in the current node (node 0). And when I further malloc new memories, they went to another node (node 1), leading poor…
user3743384
  • 91
  • 1
  • 11
0
votes
1 answer

numa_alloc_onnode() allocates memory more then it can?

I tried to use numa_alloc_onnode() to allocate 40GB of memory on a 32GB node by running the below code. However it does not report any errors. a = (int *) numa_alloc_local (sizeof(int) * GB_8 * 5); if (a == NULL) printf("a error\n"); for ( i =…
user3743384
  • 91
  • 1
  • 11
0
votes
0 answers

Where is the link between NUMA code for writing a page and the rest of Linux swap

So, for normal pages in Linux try_to_unmap creates a swap entry for a particular page and then pageout handles writing it to the swap space by calling mapping->a_ops->writepage on it. Now, shrink_page_list connects the pieces together. For NUMA…
feeling_lonely
  • 6,665
  • 4
  • 27
  • 53
0
votes
1 answer

Enabling UseNUMA flag for hadoop?

I wish to enable UseNUMA flag for my hadoop framework which uses java. The problem is I don't know which all places java command has been called by hadoop framework. So, I've written alias for java command in /etc/bash.bashrc as alias java='java…
yedapoda
  • 850
  • 2
  • 9
  • 11
0
votes
1 answer

What is a multithreading performance impact for data intensive tasks on a massive mulitcore machine?

I'm reading a post about multi-threading performance issue on massive multicore machine: http://www.reddit.com/r/Python/comments/1mn12l/what_you_do_not_like_in_python/ccbc5h8 An author of that post claims that in a massive multicore systems…
Robert Zaremba
  • 8,081
  • 7
  • 47
  • 78
1 2 3
…
20
21