Re: browser errors

[Bryan Whitehead <driver@jpl.nasa.gov>]

Korosi Akos wrote:
[snip]

You might want to run memtest to make sure you don't have a memory 
problem..... (just to rule out a mem/hardware problem)

>>Please remember that I am
>>only guessing here, but I noted your comment that "HDD light is working
>>(lighting for some time)". This makes me wonder if the system is trying to
>>use the swap partition and running into trouble (or perhaps just getting
>>VERY slow -- how long do you wait for the system to recover before
>>rebooting?). Browsers tend to be memory hogs, so you might see this
>>behavior only when running one.
> 
> 
> I put on CPU/MEM usage applet, and tried again.
> You might be right about X is not freezing, just being slow (I didn't wait for
> so many time before restart, jus about 5 minutes).
> The next happen.
> I start a browser (now everithing is ok), than log in to ISP (now ok too), 
> than start to browse the net, and then CPU usage is going mad (full).
> And everithing is slow (but not freezed now, I tried now just with opera).
> Then I started 'top' command, and saw that opera is using the most the cpu.
> And this went to 2-3 minutes, than I stopped the connection, but can not close
> opera. (It is still working, and can not be closed, now for more than 10 
> minutes, and CPU usage is full (looking at applet), and now I had to kill it 
> with -9 option).
> What can I do now?

Check you java settings, Java can do this in many different browsers. 
Kaffe (a java vm shipped with most distros) is a peice of junk, and 
might be the only java available in your machine.

Also, try using a different browser like plain netscape 4.xx or mozilla.


> 
>>If that is not it ... what are the last few things the logs contain prior
>>to a reboot out of this failure?
> 
> 
> I tried again, to generate log messages.
> Now the next happened:
> Start Opera, log to ISP, start browsing. For 1-2 minutes everything is ok.
> Then got the CPU usage to max. I wait for a while (3 minutes), the CPU usage 
> is maximal, opera is not responding to mouse click, if I drag a window over 
> opera, the window is makink a trail over opera. If I change to other 
> workspace and back, opera is white (this happened in the previous case too).
> The connection speed graph gets zigzag (little zigzag, in one second 1.5 kb 
> than 0, and again...)
> Then I disconnected from ISP, and than X aborted!
> The message of X (I start it from text mode, so I see messages after abort):
> 
> Fatal server error:
> Caught signal 11. Server aborting

This is why I ask you to check your memory settings and test the memory. 
Are you overclocking?

> (then the text that if I want to send a bug report I have to include all
>  XFree log messages...)
> 
> I could send the full XFree log, but is is around 50 kB...
> 
> And here is the /var/log/messages with messages generated at the that time:
> (I put X-es to ip addresses, because I don't know if it is a security problem 
> if I tell you. Ray wrote previously that I should cut off passwords, and I 
> don't know if I should do the same with ip-s...)
> --------------------------------------
> Aug  4 17:38:30 localhost kernel: hsf_break_ctl: modem does not presently 
> support sending/receiving BREAK
> Aug  4 17:39:04 localhost modprobe: modprobe: Can't locate module ppp0
> Aug  4 17:39:04 localhost modprobe: modprobe: Can't locate module ppp0

This means ppp is built into kernel (otherwise you'd not be able to even 
connect to your ISP at all)

> Aug  4 17:39:04 localhost pppd[11399]: pppd 2.4.1 started by root, uid 0
> Aug  4 17:39:04 localhost pppd[11399]: Using interface ppp0
> Aug  4 17:39:04 localhost pppd[11399]: Connect: ppp0 <--> /dev/ttyHSF0
> Aug  4 17:39:04 localhost /etc/hotplug/net.agent: assuming ppp0 is already up
> Aug  4 17:39:10 localhost pppd[11399]: Remote message: Login Succeeded
> Aug  4 17:39:10 localhost modprobe: modprobe: Can't locate module 
> ppp-compress-21
> Aug  4 17:39:10 localhost modprobe: modprobe: Can't locate module 
> ppp-compress-21

This might be a problem, you might not have the ppp_deflate module, or 
it's not built into the kernel. If this is true you might end up 
dropping ppp packets... But then again, we might get more error's from 
the pppd if ppp_deflate was missing.


> Aug  4 17:39:11 localhost pppd[11399]: local  IP address XXX.XXX.X.XXX
> Aug  4 17:39:11 localhost pppd[11399]: remote IP address XXX.XXX.XXX.XX
> Aug  4 17:43:20 localhost pppd[11399]: Terminating on signal 15.

This might be the most pppd will tell us about missing ppp_deflate. 
Might want to check into this... maybe the rpm didn't install fully?

> Aug  4 17:43:26 localhost pppd[11399]: Connection terminated.
> Aug  4 17:43:26 localhost pppd[11399]: Connect time 4.3 minutes.
> Aug  4 17:43:26 localhost pppd[11399]: Sent 78191 bytes, received 363790 
> bytes.
> Aug  4 17:43:26 localhost pppd[11399]: Exit.
> Aug  4 17:43:26 localhost /etc/hotplug/net.agent: NET unregister event not 
> supported

Are you using anything in a firewire or USB port?

> Aug  4 17:43:44 localhost gnome-name-server[1229]: input condition is: 0x11, 
> exiting
> Aug  4 17:43:46 localhost su(pam_unix)[11187]: session closed for user root
> Aug  4 17:44:50 localhost gconfd (korosi-1232): GConf server is not in use, 
> shutting down.
> Aug  4 17:44:50 localhost gconfd (korosi-1232): Exiting

Because X died gconfd exits as well as gnome-name-server.

-- 
Bryan Whitehead
SysAdmin - JPL - Interferometry Systems and Technology
Phone: 818 354 2903
driver@jpl.nasa.gov

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs