© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
0e1f616
(Get the code!)
(In reply to Ralf Corsepius from comment #31)
> (In reply to John Florian from comment #30)
>
> > Ralph, thanks for the excellent comment. Any news on the success (or not)
> > of using ss_en=0?
> AFAICT, in recent kernels (Don't know since when; I am on Fedora 20),
> ss_en=0 is not required anymore.
Good to know. I see the problem on my F18-based spins but not on the F20-based one nearing deployment. These two specific cases (18=bad/20=good) are both on Asus EeeBox B202 hardware, but there are many production variants of this "model" (grrrrr). Knowing you haven't seen on F20 is very helpful.
> > Anybody, if I see these messages and am without ss_en=0, does this just
> > result in log noise or does it cause ill behavior besides?
> The log noise isn't "that harmless". It had caused /var/log/messages rsp.
> journals to grow beyond reasons and caused different kinds of hick-ups on
> slow, small RAM systems (I am suspecting it to be co-responsible for hard
> machine lock ups, I once was facing with journalctl).
Hmmm... that worries me. My B202 deployments have been mostly good but more likely than other models to exhibit troublesome behavior. I should have systemd-journald conf'd where log growth isn't an issue, but ...
I actually might be able to aid my F18-based deployments after all, despite what logic and reason would have my presume. I was able to rmmod usb_realtek, make the config you offered and modprobe it back in. I had been seeing the noise about every 50s and it's now been quiet for 5m. How I was able to make the change, I don't understand. The config file only hit tmpfs (because of the statelessness), but the tools I needed to do so once rmmod'd would have had to come off SD ... or from caching of the Live squashfs.img.