DPDK usage discussions
 help / color / mirror / Atom feed
From: "Wiles, Keith" <keith.wiles@intel.com>
To: Stefano Salsano <stefano.salsano@uniroma2.it>
Cc: Muhammad Zain-ul-Abideen <zain2294@gmail.com>, users <users@dpdk.org>
Subject: Re: [dpdk-users] error running two instances of pktgen
Date: Thu, 28 Jun 2018 12:49:32 +0000	[thread overview]
Message-ID: <BCABF8DE-76B3-4E42-B75D-4A41DC27DC29@intel.com> (raw)
In-Reply-To: <d2e68ab6-31fd-a016-8681-39dfce4e6626@uniroma2.it>



> On Jun 28, 2018, at 3:53 AM, Stefano Salsano <stefano.salsano@uniroma2.it> wrote:
> 
> Il 2018-06-28 10:32, Muhammad Zain-ul-Abideen ha scritto:
>> I think it is some sort of cli command issue now, like socket-mem or blacklisting the port.
>> Adding Keith Wiles, the father of pktgen
> 
> thank you Muhammad, here is the commands that we run now and the overall recap of the problems
> 
> sudo -E ./app/x86_64-native-linuxapp-gcc/pktgen  -l 5-8 -n 3 --file-prefix pg0 --proc-type auto -b 06:00.0 -- -P -m "6.0"
> 
> it works, then we have the error in the second instance:
> 
> sudo -E ./app/x86_64-native-linuxapp-gcc/pktgen -l 0-4 -n 3 --file-prefix pg1 --proc-type secondary -b 06:00.1 -- -P -m “1.0"

Pktgen does support two or more pktgen instances but only as primary processes, which means pktgen can not be run in secondary mode so using —proc-type auto will not work.

I run pktgen as two primary processes all of the time, but I do not use the option —proc-type at all.

The problem you above looks good until you get to the ‘-m 1.0’ option. The 1 in this case denotes the lcore to use. In the above command you seemed to have corrected the problem as you were using 1.0 for both commands. Now you are using 6 and 1.

Try removing the —proc-type options and give that a try.

> 
> EAL: Detected 32 lcore(s)
> EAL: Detected 2 NUMA nodes
> EAL: Multi-process socket /var/run/dpdk/pg1/mp_socket_12381_2a519caf5992b0
> EAL: failed to send to (/var/run/dpdk/pg1/mp_socket) due to Connection refused
> EAL: Fail to send request /var/run/dpdk/pg1/mp_socket:bus_vdev_mp
> vdev_scan(): Failed to request vdev from primary
> EAL: Probing VFIO support...
> EAL: Could not map memory from primary process
> EAL: FATAL: Cannot init memory
> 
> EAL: Cannot init memory
> 
> I don't know if it is relevant, but when running the first command I get a warning about hugepages (and then the pktgen works well)
> EAL: Detected 32 lcore(s)
> EAL: Detected 2 NUMA nodes
> EAL: Auto-detected process type: PRIMARY
> EAL: Multi-process socket /var/run/dpdk/pg0/mp_socket
> EAL: No free hugepages reported in hugepages-1048576kB
> 
> 
> if I run both instances with --port-type auto it detects primary mode for both
> EAL: Detected 32 lcore(s)
> EAL: Detected 2 NUMA nodes
> EAL: Auto-detected process type: PRIMARY
> EAL: Multi-process socket /var/run/dpdk/pg1/mp_socket
> EAL: No free hugepages reported in hugepages-1048576kB
> EAL: Probing VFIO support...
> 
> and then the second instance fails like this
>    Create: Default TX  0:0  - Memory used (MBUFs 16384 x (size 2176 + Hdr 128)) + 192 =  36865 KB, headroom 128
> !PANIC!: Cannot create mbuf pool (Default TX  0:0) port 0, queue 0, nb_mbufs 16384, socket_id 0: No such file or directory
> PANIC in pktgen_mbuf_pool_create():
> Cannot create mbuf pool (Default TX  0:0) port 0, queue 0, nb_mbufs 16384, socket_id 0: No such file or directory6: [./app/x86_64-native-linuxapp-gcc/pktgen(_start+0x29) [0x466139]]
> 5: [/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0) [0x7ffff6cae830]]
> 4: [./app/x86_64-native-linuxapp-gcc/pktgen(main+0x598) [0x460388]]
> 3: [./app/x86_64-native-linuxapp-gcc/pktgen(pktgen_config_ports+0x18ad) [0x491aed]]
> 2: [./app/x86_64-native-linuxapp-gcc/pktgen(__rte_panic+0xc3) [0x45769b]]
> 1: [./app/x86_64-native-linuxapp-gcc/pktgen(rte_dump_stack+0x2b) [0x50aadb]]
> Aborted
> 
> 
> -- 
> *******************************************************************
> Stefano Salsano
> Professore Associato
> Dipartimento Ingegneria Elettronica
> Universita' di Roma Tor Vergata
> Viale Politecnico, 1 - 00133 Roma - ITALY
> 
> http://netgroup.uniroma2.it/Stefano_Salsano/
> 
> E-mail  : stefano.salsano@uniroma2.it
> Cell.   : +39 320 4307310
> Office  : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435
> *******************************************************************
> 

Regards,
Keith


  reply	other threads:[~2018-06-28 12:49 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-28  1:56 Stefano Salsano
2018-06-28  7:05 ` Muhammad Zain-ul-Abideen
2018-06-28  7:23   ` Stefano Salsano
2018-06-28  7:33     ` Muhammad Zain-ul-Abideen
2018-06-28  8:12       ` Stefano Salsano
2018-06-28  8:32         ` Muhammad Zain-ul-Abideen
2018-06-28  8:53           ` Stefano Salsano
2018-06-28 12:49             ` Wiles, Keith [this message]
2018-06-28 12:57             ` Wiles, Keith
2018-06-28  8:33         ` Muhammad Zain-ul-Abideen

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=BCABF8DE-76B3-4E42-B75D-4A41DC27DC29@intel.com \
    --to=keith.wiles@intel.com \
    --cc=stefano.salsano@uniroma2.it \
    --cc=users@dpdk.org \
    --cc=zain2294@gmail.com \
    /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).