For example, In my system with SPDK latest code in master branch, the huge page files are
mounted in /dev/hugepages, and the Huge page size is 2MB. And my SPDK process id(PID) is
53147 (this is a primary SPDK process), the huge page files used are:
spdk_pid53417map_0, spdk_pid53417map_1, spdk_pid53417map_2...
If you want to know more details, you can look at the code build_eal_cmdline function in
lib/env_dpdk/init.c file in SPDK repo.
From: Jing Liu <jingliu.cs(a)gmail.com>
Sent: Wednesday, November 18, 2020 1:57 AM
To: Storage Performance Development Kit <spdk(a)lists.01.org>
Subject: [SPDK] Re: Possible to use several hugepage files for NVMe's DMA
On Tue, Nov 17, 2020 at 8:50 AM Yang, Ziye <ziye.yang(a)intel.com> wrote:
I think that your version may be old. Currently, SPDK can support huge
page map with process id. If you use newer version, you can check the
file names in the huge page mounted folder, each primary process will
have its own huge page file. So I think that you need to update your
spdk version, e.g., spdk 20.01 LTS is recommended.
> 在 2020年11月17日，下午8:34，Jing Liu <jingliu.cs(a)gmail.com> 写道：
> In general, I'm wondering how to provide the isolation of the
> hugepage across several processes using SPDK in one computer
> In my understanding, SPDK's hugepage is initialized by the primary
> and all will use
> *shmid* to be able to find that by several processes. I'm wondering
> possible that the SPDK is initialized with several *shmid*, such
> that several processes can be isolated?
> PS: I'm using the SPDK 18.04.01, does some new version help?
> SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email
> to spdk-leave(a)lists.01.org
SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email to
SPDK mailing list -- spdk(a)lists.01.org
To unsubscribe send an email to spdk-leave(a)lists.01.org