Date: Wed, 14 Feb 2018 11:00:59 -0600 (CST) From: "Valeri Galtsev" <galtsev@kicp.uchicago.edu> To: "Polytropon" <freebsd@edvax.de> Cc: "Valeri Galtsev" <galtsev@kicp.uchicago.edu>, "FreeBSD Mailing List" <freebsd-questions@freebsd.org> Subject: Re: OT: just crashed two FreeBSD-11.1-RELEASE machines Message-ID: <33865.128.135.52.35.1518627659.squirrel@cosmo.uchicago.edu> In-Reply-To: <20180214011333.23d616b9.freebsd@edvax.de> References: <cf4680fd-32fc-d22c-602d-b5445db51549@kicp.uchicago.edu> <20180214011333.23d616b9.freebsd@edvax.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, February 13, 2018 6:13 pm, Polytropon wrote: > On Tue, 13 Feb 2018 16:56:12 -0600, Valeri Galtsev wrote: >> I just crushed two (one at a time) FreeBSD-11.1-RELEASE machines. I >> simply plugged my Sony video camera which was in the mode of exporting >> its (msdos) filesystems over USB. Both machines have decent hardware, >> plenty of resources. Regular user was logged in to X11. One of the >> filesystems was mounted (the one holding AVCHD videos), and two seconds >> later the machine started booting (like "crashed"). >> >> And now I'm sitting in prostration and thinking: What is that I will do >> next? > > Gather debugging data. Reduce automatisms and check each step > manually. Try to find out the exact "thing" that brought down > the system (it's hard to tell when X is running and doing its > "magic"). > > Many years ago, I had a similar problem, but when _detaching_ > the camera from the system (internal filesystem was already > unmounted, and previous access to it went as expected)... > This was, I think, in FreeBSD 5 times, and after a system > update, the problem went away. So I never found out what the > actual problem was - but it could be reproduced any time, > even in single user mode: insert USB plug, system recognizes > da0 device (as intended), wait a few seconds, remove USB plug, > system panic, kernel crash. After update: regular "disconnect" > message, nothing else, system kept running. > Yes, indeed, my duty as a user is to keep reproducing it, collect debugging information and pass that over to developers so they can improve the system. My problem, however is of different nature. I did migrate a bunch of servers to FreeBSD. And I still somehow trust that these servers will be as reliable as they were so far (which is 2 or 3 years). However, I am about to give the developer FreeBSD laptop, thus the person (who develops stuff that runs on servers of one's own laptop) can have development environment the same as that on production servers. And I can not do it now, knowing that people (myself included) plug into laptops all and everything... that can potentially knock down FreeBSD box just like I just did. This whole situation gives me deja vu. Long ago when I used RedHat mostly I gave a try to a SUSE Linux. That was somewhere near SUSE 7.1-7.3. The main trouble with SUSE was: if I start as a regular user something with memory leak (I just had trivial test code that keeps allocating memory in the loop), then SUSE box crashed. If, however I just build stock kernel from kernel.org, it happily invokes OOM killer, and kills the offender. Needless to say, I never used SUSE, and I never seriously looked back at SUSE. And now I have this deja vu... Thanks for helping my mood, which is still way from good... Valeri ++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?33865.128.135.52.35.1518627659.squirrel>