Delayed booting problem on F7
David Timms
dtimms at iinet.net.au
Sat Oct 13 00:40:20 UTC 2007
Gordon Messmer wrote:
> Karl Larsen wrote:
>> This morning I started the computer and it stopped for 10 minutes
>> because it could not find cups. It talked about applying iptables but
>> had "never matched protocal" and when it finally came up Thunderbird
>> was broken.
>
...
> After checking your SELinux settings, you should have checked resolution
> of your hostname. To do this, start by establishing what your hostname is:
>
> # hostname
> herald.private.dragonsdawn.net
>
> Then make sure that hostname can be resolved, either by the hosts file
> or DNS:
>
> # getent hosts `hostname `
> 192.168.1.6 herald.private.dragonsdawn.net
>
> If you get an error from "getent", then you need to make sure that your
> hostname can be resolved. You can do this by listing it in /etc/hosts,
> or by setting up the name in DNS.
This is my suspicion for Karl's issue as well. Even though it isn't my
issue, I would like to confirm the root cause of the problem.
A final test is does ping `hostname` actually get "reply"s ?
ie: ping k5di
Even if it the hostname is pingable now, I think it would not have been
pingable at the time when Karl had the initial problem. His contents of
/etc/resolv.conf showed the comment from dhclient. I'll put down some
further possibilities that may have led to this issue {that seemed to
resolve itself}.
Since likely a DSL connection to the internet:
- Network cable faulty or partially plugged in {wired} ?
- Assuming a DSL router does the pppoe login:
- powered off
- rebooting at the inopportune moment {check the uptime of the router}
- relogging on due to inactivity timeout {check the router logs - and
that it time marks are correct time}
- failed to give out an dhcp address ?
- isp dns server down at the time the pc is trying to resolve its
hostname {instead of an instant ~"does not exist", PC's dns gets no
answer, timing out after 1 minute, then trying again a couple of
times...} {check the ISPs service status - decent ones tell you when
they have had issues} {also causes slow startup for thunderbird since it
can't resolve the mail server name}.
- the received dhcp address has a domain that does not exist {if I
set domain in my router's dhcp server as "zzz", then for eg sendmail
needs to be able to resolve `hostname`.zzz - it will timeout after 1
minute, try again, time out after the second minute, then sm-client will
also timeout - leads to a 3 minute boot delay}.
- A change to a PC setting:
- hostname changed {and not updated in /etc/hosts} ?
- firewall config change ?
- package update {less /var/log/yum.log} ?
eg: Sep 28 22:47:50 Updated: iptables-ipv6 - 1.3.8-2.1.fc7.i386
eg: Oct 11 22:54:16 Updated: dbus-glib - 0.73-3.fc7.i386
{it might help us to see the most recent 50 entries from yum.log}
- An issue in a package
- network manager, dbus, dhclient, iptables ?
- "fail to find cups and then sendmail":
- The description is poor. A better one might be "the system shows:
Starting cups ... and then sits there for some minutes. This also
occurred at the starting sendmail / sm-client ... part". It is not that
the system can't find either of these programs, but that they are taking
more than normal time to startup.
- "Thunderbird was broken"
- There was no further mention of the nature of the issue with
thunderbird. Was it delay in getting email ? Were there any error
messages ? what exactly was the issue ? Was it delay in starting
thunderbird at all ?
I have found it useful to have a pen/paper near my computer so that I
can catch any weirdness as close to the letter as possible. I also have
an old digital watch there, so that I can time {in seconds} something
that seems to be taking to long - sometimes that is enough to indicate
the type of fault to others.
I am sure there are more possibilities that can lead to the symptoms
described by Karl.
DaveT.
More information about the fedora-list
mailing list