|
Bugzilla – Full Text Bug Listing |
| Summary: | dhcpcd clobbers interface settings | ||
|---|---|---|---|
| Product: | [openSUSE] SUSE Linux 10.1 | Reporter: | Andreas Schwab <schwab> |
| Component: | Network | Assignee: | Peter Poeml <poeml> |
| Status: | RESOLVED WORKSFORME | QA Contact: | E-mail List <qa-bugs> |
| Severity: | Major | ||
| Priority: | P5 - None | ||
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PowerPC | ||
| OS: | Other | ||
| Whiteboard: | |||
| Found By: | Other | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
|
Description
Andreas Schwab
2005-11-12 10:10:28 UTC
Do you have more details? How do you know that dhcpcd is the culprit? There is no evidence that anything else should be it. Once the address got lost, is dhcpcd running then? what does /var/lib/dhcpcd/dhcpcd-eth0.info contain? what does grep dhc /var/log/messages show, after enabling debugging in /etc/sysconfig/network/dhcp? Nov 12 10:58:15 whitebox modify_resolvconf: restored /etc/named.conf.saved.by.dhcpcd.wlan0 to /etc/named.conf Nov 12 10:58:25 whitebox modify_resolvconf: Service dhcpcd modified /etc/named.conf. See info block in this file This rather looks as if dhcpcd takes down the interface (the lease might have become invalid), and has acquire a new lease after 10 seconds, so it rewrites named.conf again. The debug log would help. The lease time is how long? LEASETIME=86400 RENEWALTIME=43200 REBINDTIME=75600 -rw------- 1 root root 136 Jan 8 12:14 dhcpcd-wlan0.cache -rw-r--r-- 1 root root 518 Jan 8 12:14 dhcpcd-wlan0.info -rw-r--r-- 1 root root 518 Jan 3 13:12 dhcpcd-wlan0.info.old -rw-r--r-- 1 root root 0 Jan 3 13:12 dhcpcd-wlan0.timestamp => Might be related to bug 131295 Puzzling... How often does it appear? - every 12 hours - every 21 hours - every 24 hours - irregularly The latter. Status of the bug? Sorry about the late answer. I assume that this is still reproducable? If so, could you please grab an strace log with timestamps, like strace -ftt -o dhcpcd.strace -p $(pidof dhcpcd) from when it happens? If it happens irregularly, I have no idea what could be causing it. As I wrote above, I hope that switching debug mode on does reveal something. It can be enabled by setting DHCLIENT_DEBUG in /etc/sysconfig/network/dhcp to "yes". However, I rather suspect a problem with the driver; maybe the device becomes unavailable at times... which network driver is it that you are using? This happened with the airport driver. In the G3 Powerbook, the driver works fine (I use it all the time). Even temporary link losses don't cause such a problem here: Jun 16 15:41:17 cherry kernel: eth1: New link status: AP Out of Range (0004) Jun 16 15:41:17 cherry kernel: eth1: New link status: AP In Range (0005) Jun 16 15:41:51 cherry kernel: eth1: New link status: AP Out of Range (0004) Jun 16 15:41:51 cherry kernel: eth1: New link status: AP In Range (0005) ...dhcpcd continues to work. Hi Andreas, does the problem still exist with your machine? This machine does not exist any more. Okay. Closing this bug then. |