xcdroast causing network connection to timeout?!?

xcdroast causing network connection to timeout?!?

Post by Peter Edd » Wed, 10 Mar 1999 04:00:00

This is a weird one, in my opinion.  I've been happily using xcdroast
for a while now and haven't had any problems.  Today I was online and
thought I'd begin reading an audio CD to disk in order to copy it.

As soon as xcdroast started reading the audio tracks, my network
connection became unuseable.  At first I thought it was just the site I
was looking at, so I tried other sites, nothing.  I tried FTPing (with
ncftp), couldn't connect to any servers.  I cancelled the xcdroast read
and the 3 netscape windows that were stalled came to life and loaded
their pages in a second or two.  I then continued to browse and all was
well.  Then I restarted the read and the netconnection became
unresponsive again.  I cancelled the read and everything came back to
life.  At all times, xload reports that the CPU is very low, approaching
nothing and the system 'feels' like there's no load too.

What's going on?  Anyone else experienced this?

Here's my machine info:

RedHat 5.2, kernel 2.0.36 w/IDE SCSI emulation
Buslogic BT
128M Ram
Celeron/400 (not overclocked)
3Com ImpactIQ ISDN TA (serial)

xcdroast 0.96e w/the cdparanoia patch (cause cdrecord didn't work right
at the time)
cdparanoia III  alpha prerelease 9 (october 9, 1998)



1. Slow Internet connection or sluggishness cause Telnet session timeouts


This is an update to my last post, which, -unless I've missed the
reply-, I have not received any answer.
I have a couple "legal" class C address which I'm trying to reduce to a
bare minimum through IP Masquerading.

I currently have a dual NIC Linux Masquerade server, "eth0" with a
"real" class C address, "eth1" as the local class B ,
with local DNS, and another Linux box with legitimate class C as an
external DNS server and sendmail server.

I use an NT machine for DHCP and WINS server. Works beautifully, for the
NT environment, although I have
experienced very infrequent lost connection to the NT file server. After
a long period of inactivity,
sometimes, I would get the message " Server xxxxxx is no longer
accessible". Browsing the network
neighborhood will invariably get it back. Internet access is faster than

The main problem resides with Telnet and Internet access.

I have evidence now that my Telnet sessions to the local "real class C"
systems, be it Vaxes,
Alphas, Suns or other Unix systems, including the Linux Masquerade
server, will disconnect and drop out
when I encounter a slow internet connection ??? Typically, when
accessing busy sites, while the browser
is trying to make a connection for more than a minute or two, the Telnet
sessions will be disconnected.
Occasionally it will totally freeze the Win95 machine, and the only
recourse is to reboot.

I am beta testing a dozen users within different groups on this new
internal Class B( 172.16.x.x). They all experience the same symptoms.
The minute I revert back to the legitimate IP address, all problems went
away. All seems to point to the masquerading environment. I was under
the impression that IP masquerading timeout default value is 15 minutes,
although I'm not sure how to
check it. I assure you that after all those disconnect, I never leave a
connection attempt on for more than 60 sec. My ipforwarding setup is
very simple for now, just couple lines of deny everyone then allow
forwarding from Internal to External.
Has anyone encountered similar anomaly?

I'm new to Linux, and would appreciate all your feedback, help and

Much obliged,

Gerard Daniel

2. getting PSmon working

3. A Strange connection problem with no reasonable cause on network

4. Linux

5. newbie: mac address causing network connections to fail

6. Replace pcihpfs with sysfs.

7. HOWTO Timeout idle network connections

8. Xircom Creditcard ethernet (PCMCIA)

9. Cannot accept chooser connection Software caused connection abort

10. Anwser: Possible causes to crashes caused by network (3c509)

11. Unexpected connection timeouts on local TCP connections

12. FreeNX: seession resume caused timeout

13. hda: timeout while waiting for DMA causes crahses (incessantly)