Jonathan Perkin · @jperkin
302 followers · 240 posts · Server federate.me.uk

It's a pity that Mastodon lacks quote tweets, but in addition to the post I boosted yesterday, I just wanted to say my own thanks for reaching 20.

I use it all the time, and struggle when working on other systems that do not have it. Only yesterday I wrote a quick script to verify a code path that lead to this fix: github.com/NetBSD/pkgsrc/commi.

It has found significant performance wins in and , fixed numerous bugs, and helped me understand systems.

Thanks @bcantrill and @ahl!

#dtrace #pkgsrc #pkgin

Last updated 1 year ago

Jonathan Perkin · @jperkin
297 followers · 224 posts · Server federate.me.uk

As part of my recent work I came across another optimisation for "pkg_admin rebuild-tree" using that makes it a further 12x faster on my test system.

mail-index.netbsd.org/tech-pkg

This is in addition to the 13x speedup I committed a few years ago:

gist.github.com/jperkin/98550d

Reviews appreciated.

#pkgin #dtrace #pkgsrc #netbsd

Last updated 1 year ago

Jonathan Perkin · @jperkin
285 followers · 192 posts · Server federate.me.uk

After some more analysis and code reading, found another easy optimisation win in the pattern matching code, for a further 2x speedup.

Next up some database optimisations which will hopefully get us to around 8x combined speedup, as well as helping with future development work.

#dtrace #pkgsrc #pkgin

Last updated 1 year ago

Dave Marquardt · @davemq
61 followers · 420 posts · Server fosstodon.org

Today I used on a system to investigate a performance issue. Really took me back to my days working on at Sun Microsystems. I used quite a bit at that time, and it's good to see the same good ideas thriving in Linux.

I worked in development at for many years, and I was always a little sad that people didn't use more. Perhaps this is because AIX has a good system trace infrastructure and the culture grew up using that. Hard to say.

#bpftrace #linux #solaris #dtrace #aix #ibm #probevue

Last updated 1 year ago

Jonathan Perkin · @jperkin
283 followers · 189 posts · Server federate.me.uk

After maintaining for a few years, there's still a bunch of low-hanging fruit for cleanup and optimisation.

While working on SUPERSEDES support, I was curious as to why the dependency calculation takes so long.

Re-ran it with , saw break_depends() taking up 55% of the runtime. What does it do?

Turns out, nothing. Various changes over the years, notably refresh support, have made its calculations obsolete.

Nice easy 2x performance win and much easier-to-understand code path.

#pkgin #dtrace

Last updated 1 year ago

Viraptor · @viraptor
91 followers · 270 posts · Server cyberplace.social

Ok, another person confirmed it. on max with Ventura is completely broken.
Feel free to mention existing feedback FB12061147 if someone ends up reporting it as well.

#dtrace #m2 #mbp

Last updated 2 years ago

Viraptor · @viraptor
89 followers · 268 posts · Server cyberplace.social

I give up. just doesn't work on anymore, does it? The most basic invitations freeze my machine every single time 😒

#dtrace #m2 #mbp

Last updated 2 years ago

Viraptor · @viraptor
90 followers · 225 posts · Server cyberplace.social

Grrr.... : Hey, we've got this awesome available. Also MacOS: lol, write your own tools, even dtruss is broken, can't trace forks, and we'll print fd numbers in decimal in return values and hex in arguments - screw you devs.

#macos #dtrace

Last updated 2 years ago

Mina · @meena
432 followers · 10789 posts · Server cathode.church

the fact that this laptop is trying to fly away because one tiny virtual machine is hitting one virtual CPU at 100% is beyond ridiculous

… and the reason I'm contemplating diving into the manuals at 01:55 in the mor…night.

because, like, doesn't this dtrace_kinst(4) look really easy??

#dtrace

Last updated 2 years ago

Pierre Lebeaupin · @PierreLebeaupin
11 followers · 114 posts · Server mastodon.gougere.fr

Update: I *finally* located the post that to me best summarizes the philosophy in question: queue.acm.org/blogposting.cfm? (this is the ACM repost, appropriately enough). Cc @bcantrill @ahl

#gdb #gdbserver #dtrace #unix #socat #ssh #eclipse #linux

Last updated 2 years ago

Jeremy 🇨🇦 · @thatgeoguy
73 followers · 588 posts · Server coales.co

Pleasantly surprised by "BPF Performance Tools" by Brendan Gregg: It's already changed how I'm thinking about debugging and profiling across my system.

I guess I should have listened to all those folks telling me to learn and for so long.

#dtrace #bpftrace

Last updated 2 years ago

· @lamitpObuS
32 followers · 276 posts · Server troet.cafe

@lattera Could be an acceptable replacement? (I don't know and have only limited experience with bpftrace)

brendangregg.com/blog/2018-10-

#dtrace #bpftrace

Last updated 2 years ago

Shawn Webb · @lattera
1419 followers · 5301 posts · Server bsd.network

I really wish had . It would make security auditing of
proprietary applications so much easier.

#linux #dtrace #winwithbsd #runbsd

Last updated 2 years ago

Alexandre Borges · @alexandreborges
188 followers · 12 posts · Server infosec.exchange

DTrace is great for instrumenting kernel, drivers, malware samples and vulnerability research. I will use it in one of the next articles.

#windows #dtrace

Last updated 2 years ago

Jonathan Perkin · @jperkin
205 followers · 89 posts · Server federate.me.uk

@sjmulder You might want different graphs for different timing data, e.g. runtime vs CPU time, on you can get these out of curpsinfo on process exit but while I think curpsinfo itself is available on all OS the members themselves may vary.

As a quick hack for printing the process tree on illumos I'd system() a ptree(1) on the exiting process and then filter on interesting stacks.

Experimenting with is fun! Would recommend just playing and see what you come up with.

#illumos #dtrace

Last updated 2 years ago

Jonathan Perkin · @jperkin
205 followers · 89 posts · Server federate.me.uk

@sjmulder You probably want the output from the script to be similar to stack trace output, where you have the full process tree for each leaf process, and at the bottom the time spent in that leaf process.

#dtrace

Last updated 2 years ago

Jonathan Perkin · @jperkin
205 followers · 89 posts · Server federate.me.uk

@sjmulder The answer is , start with something like execsnoop but keep a track of parent/child relationships and how long the processes ran for, then use the tools in the repository to convert the text output to an SVG.

#dtrace #flamegraph

Last updated 2 years ago

Jonathan Perkin · @jperkin
177 followers · 52 posts · Server federate.me.uk

Sometimes performance work is little more than watching execsnoop for duplicate lines while builds are running.

Spotted a huge number of chmod processes being spawned, very quickly tracked it down to "find -exec" using "{} \;" instead of "+".

Tiny change, but reduces number of execs by over 25,000 and significantly reduces build and system time.

If you're interested in hacking on there is plenty of very rewarding, very easy to fix, low-hanging fruit like this!

#dtrace #pkgsrc

Last updated 2 years ago

Luca De Pandis · @ldep
12 followers · 63 posts · Server bsd.network

So things are getting bad, and is being considered more complex to manage and more costly.
I have to admit that while I really believed in the cloud, the problems that my customers were facing started to break my beliefs.
Ops costs increased, because they still hired the IT guy, but now a costs way more than a traditional , and often he has an knowledge way inferior, since everything is not in house anymore. Everything is hosted and you don't access to the technology internals anymore.
Also, they are dealing with a level of complexity unheard of.
Before the cloud there were only physical and virtualized layers of complexity.
Now we also have then and layers.
We moved from putting a stupid class in classpath to run with the profiling to running , , and to obtain the same level of informations.

The only good thing is that now in Linux engineers are using to profiling network and software execution.
But this is NOT innovation at all. You could do that already with in and with way less complexity, and the same level of profiling since , and lot of frameworks included its probes to support advanced tracing.
Apple's used to provide a good frontend called that was really cool to work with.

world.hey.com/dhh/why-we-re-le

#cloud #SiteReliabilityEngineer #systemadministrator #infra #containerorchestration #servicemesh #java #jboss #apm #prometheus #grafana #jaeger #kiali #ebpf #dtrace #solaris #freebsd #ruby #macos #instruments

Last updated 2 years ago