Red Hat Bugzilla – Bug 52285
PPPD dies with error 53 or 9 or 5
Last modified: 2007-04-18 12:36:10 EDT
From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 5.0; Windows 98; DigExt) Description of problem: pppd dies ever so often with a error 53 ect when I try to connect to my isp with a ifup ppp0 command Sometimes it also hangs the shell to such a extent that I have to kill the shell get rid of it Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.configure ppp interface from kde 2.type ifup ppp0 at prompt or in shell script 3. Actual Results: sometimes the shell will hang or pppd will report a error 53 or 9 or 5 Expected Results: the shell sould have returned to the prompt or the rest of the shell script should have executed Additional info: the script that I am trying to use are as follows #!/bin/bash /sbin/ifup ppp0 sleep 5 fetchmail -d0 -v --nosyslog sleep 5 /usr/sbin/sendmail -q sleep 2 /sbin/ifdown ppp0
I just updated from 7.0 to 7.1 (with latest updates) and i'm getting the same thing. There quite a few bug reports that also look like they are related...going back to 2-2001 (bugzilla search ppp and initscripts) What appears to happen is that ppp-watch is hanging and complaining about not finding /var/run/ppp-ppp0.pid and other goofyness. I downloaded the source for both initscripts and ppp so I could look at the source and see what going on and to find out what the errpr codes were. It will dial, connect and then kill itself. One of the workarounds includes editing 'cfgup-ppp' and commenting out # exec /sbin/ppp-watch "${DEVICE}" "$@" ..and then letting the exec pppd line at the bottom fire it up. This also requires an edit which is to remove the leading '-' from the updetach option #exec /usr/sbin/pppd -updetach $opts ${MODEMPORT} ${LINESPEED} \ exec /usr/sbin/pppd $opts ${MODEMPORT} ${LINESPEED} \ ipparam ${DEVNAME} linkname ${DEVNAME} call ${DEVNAME}\ noauth \ ${PPPOPTIONS} which resolves the "/usr/sbin/pppd: unrecognized option '-updetach'" problem. This does not correct the fact you can no longer run ppp-watch having a reconnect on failure setup or bring up ppp0 at boot time This machine had a working setup that I had been using for years. I moved my files aside and let rp3 create new ones and the problem persisted. If I uncomment the 'exec /sbin/ppp-watch' line now that I have it rigged it will it will go back to hanging/failing. For now I have a cron looking for pppd and restarting it. #!/bin/ksh /bin/ps cuax | /bin/grep ' pppd' result="$?" if [ $result == "1" ]; then print "Not found..starting pppd"; /etc/sysconfig/network-scripts/ifup-ppp ppp0 & else print "Started pppd" fi ---------------- Version-Release number of selected component (if applicable): 7.1 with all updates initscripts-5.84-1 ppp-2.4.0-2----------------
I made a mistake in my report The default line in ipup-ppp is: exec /usr/sbin/pppd -detach $opts ${MODEMPORT} ${LINESPEED} \ ipparam ${DEVNAME} linkname ${DEVNAME} call ${DEVNAME}\ noauth \ ${PPPOPTIONS} I changed '-detach' to 'updetach': exec /usr/sbin/pppd updetach $opts ${MODEMPORT} ${LINESPEED} \ ipparam ${DEVNAME} linkname ${DEVNAME} call ${DEVNAME}\ noauth \ ${PPPOPTIONS} ...in order for it to detach and work properly for me while having the 'exec /sbin/ppp-watch' line commented out. Otherwise 'ifup-ppp ppp0' hangs and ctrl-c kills the connection. Sorry about that.
Please verify this with a newer version of Red Hat Enterprise Linux or Fedora Core and reopen it against the new version if it still occurs. Closing as "not a bug" for now.