DPDK usage discussions
 help / color / mirror / Atom feed
From: "Wiles, Keith" <keith.wiles@intel.com>
To: Kumaraparameshwaran Rathnavel <krath@cloudsimple.com>
Cc: Kyle Larose <klarose@sandvine.com>, "users@dpdk.org" <users@dpdk.org>
Subject: Re: [dpdk-users] Failed to remap 2 MB pages
Date: Wed, 2 Aug 2017 13:31:57 +0000	[thread overview]
Message-ID: <AE77B894-5698-4BC4-B4B4-7C507A598FC9@intel.com> (raw)
In-Reply-To: <6882F2EB-D0B1-4627-844A-E284883A8CD9@cloudsimple.com>


> On Aug 2, 2017, at 2:01 AM, Kumaraparameshwaran Rathnavel <krath@cloudsimple.com> wrote:
> 
> Hi Kyle,
> 
> I have rebooted it couple of times. Did not have any impact, I doubt if that is the case because in the code before the mmap call there is check for the memory.
> 
> Please correct me if I am wrong.

Do you setup your huge pages via the sysctl.conf file or by hand after you startup the system?

One other item is I had something similar in my dual socket machine. I could only get about 12K total 2megs pages in 64G machine 32G per socket. I increase the memory per socket and that allowed me to reach 16K pages total or 8K pages per socket. I did not understand why it happened I just assumed it was how Linux used memory. Sorry this most likely does not help much.

> 
> Thanking You,
> Param.
>> On 02-Aug-2017, at 11:34 AM, Kyle Larose <klarose@sandvine.com> wrote:
>> 
>> Could it have to do with memory fragmentation? What if you reboot? What if you configure the huge pages on the grub command line so they are allocated at boot?
>> 
>> -----Original Message-----
>> From: users [mailto:users-bounces@dpdk.org] On Behalf Of Kumaraparameshwaran Rathnavel
>> Sent: Tuesday, August 01, 2017 7:41 PM
>> To: users@dpdk.org; Wiles, Keith
>> Subject: [dpdk-users] Failed to remap 2 MB pages
>> 
>> Hi All,
>> 
>> I am getting the error Failed to remap 2 MB pages when I give 32GB of memory to DPDK. I have a 256GB system and it was working for 16GB and I had 8192 number of Huge pages previously. Now for 32GB I have increased it to 16384 entries and I am facing this error. I guess this is not because of the unavailability of physical memory. 
>> 
>> Please correct me if I am wrong,
>> 
>> Thanking You,
>> Param.
> 

Regards,
Keith

      reply	other threads:[~2017-08-02 13:32 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-08-01 14:11 Kumaraparameshwaran Rathnavel
2017-08-02  6:04 ` Kyle Larose
2017-08-02  7:01   ` Kumaraparameshwaran Rathnavel
2017-08-02 13:31     ` Wiles, Keith [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=AE77B894-5698-4BC4-B4B4-7C507A598FC9@intel.com \
    --to=keith.wiles@intel.com \
    --cc=klarose@sandvine.com \
    --cc=krath@cloudsimple.com \
    --cc=users@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).