From: Trevor Highland <tsh@vectra.ai>
To: "users@dpdk.org" <users@dpdk.org>
Subject: [dpdk-users] Ixgbe empty queue stats
Date: Wed, 24 Feb 2021 13:24:37 +0000 [thread overview]
Message-ID: <DM5PR08MB3434A53C75B180B834951DCFA49F9@DM5PR08MB3434.namprd08.prod.outlook.com> (raw)
Hi all,
We have a DPDK application that’s running using 82599 NICs with the ixgbe driver. Our application makes use of RSS and runs with 8 or 16 queues depending on the environment. I was reviewing xstats from a DPDK application that we run, and found that ‘rx_qN_packets’ was 0 for all queues except queue 0.
I was able to reproduce this using dpdk-testpmd from the 20.11 release. Additionally I used the ethregs tool and confirmed that the registers are indeed 0. Is this expected behavior, and is there a proper way to get per queue packet counts from hardware registers?
./dpdk-testpmd -l 0-8 -n 4 -m5120,5120 -- -i --forward-mode=rxonly --rxq=2 --txq=2 --portmask=0xc --rxd=65536 --nb-cores=8 --rss-ip
I then run ‘start’
Truncated uput from ‘show port xstats all’
###### NIC extended statistics for port 2
rx_good_packets: 480147
tx_good_packets: 0
rx_good_bytes: 558567639
tx_good_bytes: 0
rx_missed_errors: 602293
rx_errors: 0
tx_errors: 0
rx_mbuf_allocation_errors: 0
rx_q0_packets: 480147
rx_q0_bytes: 558567639
rx_q0_errors: 0
rx_q1_packets: 0
###### NIC extended statistics for port 3
rx_good_packets: 419519
tx_good_packets: 0
rx_good_bytes: 438407091
tx_good_bytes: 0
rx_missed_errors: 441020
rx_errors: 0
tx_errors: 0
rx_mbuf_allocation_errors: 0
rx_q0_packets: 419519
rx_q0_bytes: 438407091
rx_q0_errors: 0
rx_q1_packets: 0
rx_q1_bytes: 0
rx_q1_errors: 0
Output at shutdown which shows both queues servicing packets
------- Forward Stats for RX Port= 2/Queue= 0 -> TX Port= 3/Queue= 0 -------
RX-packets: 116794 TX-packets: 0 TX-dropped: 0
------- Forward Stats for RX Port= 3/Queue= 0 -> TX Port= 2/Queue= 0 -------
RX-packets: 156055 TX-packets: 0 TX-dropped: 0
------- Forward Stats for RX Port= 2/Queue= 1 -> TX Port= 3/Queue= 1 -------
RX-packets: 607297 TX-packets: 0 TX-dropped: 0
------- Forward Stats for RX Port= 3/Queue= 1 -> TX Port= 2/Queue= 1 -------
RX-packets: 451971 TX-packets: 0 TX-dropped: 0
---------------------- Forward statistics for port 2 ----------------------
RX-packets: 724010 RX-dropped: 0 RX-total: 724010
TX-packets: 0 TX-dropped: 0 TX-total: 0
----------------------------------------------------------------------------
---------------------- Forward statistics for port 3 ----------------------
RX-packets: 607924 RX-dropped: 0 RX-total: 607924
TX-packets: 0 TX-dropped: 0 TX-total: 0
----------------------------------------------------------------------------
+++++++++++++++ Accumulated forward statistics for all ports+++++++++++++++
RX-packets: 1331934 RX-dropped: 0 RX-total: 1331934
TX-packets: 0 TX-dropped: 0 TX-total: 0
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Thanks,
Trevor
next reply other threads:[~2021-02-27 7:54 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-02-24 13:24 Trevor Highland [this message]
2021-02-27 18:45 ` Stephen Hemminger
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=DM5PR08MB3434A53C75B180B834951DCFA49F9@DM5PR08MB3434.namprd08.prod.outlook.com \
--to=tsh@vectra.ai \
--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).