[RFC PATCH 0/7] Add script for real-time telemetry monitoring
Stephen Hemminger
stephen at networkplumber.org
Sat Dec 13 00:23:11 CET 2025
On Fri, 12 Dec 2025 17:52:15 +0000
Bruce Richardson <bruce.richardson at intel.com> wrote:
> On Fri, Dec 12, 2025 at 02:32:49PM +0900, Stephen Hemminger wrote:
> > On Wed, 10 Dec 2025 16:55:25 +0000
> > Bruce Richardson <bruce.richardson at intel.com> wrote:
> >
> > > TL;DR
> > > ------
> > >
> > > For a quick demo, apply patces, run e.g. testpmd and then in a separate
> > > terminal run:
> > >
> > > ./usertools/dpdk-telemetry-watcher.py -d1T eth.tx
> > >
> > > Output, updated once per second, will be traffic rate per port e.g.:
> > >
> > > Connected to application: "dpdk-testpmd"
> > > Time /ethdev/stats,0.opackets /ethdev/stats,1.opackets Total
> > > 16:29:12 5,213,119 5,214,304 10,427,423
> > >
> > >
> <snip>
> > >
> > > usertools/dpdk-telemetry-watcher.py | 429 ++++++++++++++++++++++++++++
> > > usertools/meson.build | 1 +
> > > 2 files changed, 430 insertions(+)
> > > create mode 100755 usertools/dpdk-telemetry-watcher.py
> > >
> > > --
> > > 2.51.0
> >
> >
> > The TUI patch was
> > https://patchwork.dpdk.org/project/dpdk/patch/20220831115250.362189-2-conor.walsh@intel.com/
> >
>
> Took a look at that patch proposal, and here are some of the main
> differences between the two scripts:
>
> * the telemetry-tui is a full-screen app which is a lot prettier and
> probably much more suitable for non-experts.
> * the script I provided is FAR less user friendly and operates at a much
> more "techie" level
> * on the other hand, it's more flexible because of that:
> - can be used to display any numeric stats - the tui is hard-coded to show
> particular NIC stats - including e.g. cryptodev, or dmadev ones
> - can show deltas and totals as desired
> - can quit after a certain fixed amount of time
> * my script is probably more limited in terms of what can be displayed,
> since it assumes that you want one-line of output per update. TUI
> displays lots of (fixed)stats simultaneously
> * TUI as proposed has a few glitches - got a divide by zero error when run
> against a testpmd instance that didn't have traffic running, so all stats
> were zero. (That's an easy fix though)
> * The new script I propose has one major feature that I would look to add
> to the TUI if we merged that - the reconnection support. I find it very
> handy to run the watcher script and leave it running while I start and
> stop whatever app(s) I'm using.
> * My script has no additional dependencies beyond stock python, while the
> TUI requires one additional dependency "rich" to manage the TUI. It also
> can use a second dependency "plotext" (not Ubuntu packaged) to display
> charts if so desired.
>
> Just my quick assessment of the two from a quick try of the TUI one.
>
> /Bruce
It is not a case of A or B. Both tools are useful.
I can see uses for both. But original submission of TUI seems to have
been a one off patch with no long term followup
More information about the dev
mailing list