I've said it before and I'll say it again. If you have a cable modem do
not use pump. Use dhcpcd. If you are using pump and there is no problem,
it is because you have an extraordinarily long lease time (maybe -1).
I'm on RCN and pump is a disaster. It is not capable of starting a new
lease in the event of the old lease failing to renew. It will abort.
The change to be made to convert over to dhcpcd is in /sbin/ifup.
I am not aware of anything to indicate that this so-called non-busted pump
is capable of starting a new lease if the old one fails to renew.
--
-Time flies like the wind. Fruit flies like a banana. Stranger things have -
-happened but none stranger than this. Does your driver's license say Organ
-Donor?Black holes are where God divided by zero. Listen to me! We are all-
-individuals! What if this weren't a hypothetical question? [EMAIL PROTECTED]
On Sat, 14 Oct 2000, Michael H. Warfield wrote:
=>On Sat, Oct 14, 2000 at 09:50:33PM -0400, Charles Galpin wrote:
=>> On a rh6.2 system I use pump-0.7.8-1 with a dsl/dynamic ip connection, and
=>> this sort of thing works for me. You probably need to specify the
=>> interface - 'pump -i eth0 -s'
=>
=>> Mike, Hal is right. If you use the ability to run a script when the
=>> interface goes up or down, or the lease renews you could dump this info to
=>> a file yourself. If it's the format you don't like, send me an example of
=>> the format you would like to see, and I'll make a script to generate that
=>> for you if you'd like. It sounds useful, although I have found that just
=>> the IP is all I need (which is already provided by pump with having to do
=>> any scripting)
=>
=> Trust me on this one... You would probably have a tough time
=>out scripting me (in any one of SEVERAL languages and shells). :-)
=>
=> On that note, however... What advantages does pump (the current
=>non-BUSTED one, assuming it's stable) have over dhcpcd? It would SEEM
=>to do very little (the scripting capability is actually a little better
=>with the qualifications) that dhcpcd does not do. Why bother writing
=>a script when a program already exists which does the job for you in a
=>neat canned AND STABLE configuration? Pump has changed a lot over the
=>last few revs. I'm not sure I trust it. Dhcpcd has been pretty stable.
=>There is only one thing that is giving me troubles, and it's a problem
=>with both... I need a way to prohibit them from replacing my default
=>route. Maybe the dhcpclient from the ISC dhcp package will do that for
=>me as well... I'm still looking into a lot of things. ;-/
=>
=>> charles
=>>
=>> On Sat, 14 Oct 2000, Hal Burgiss wrote:
=>>
=>> > On Sat, Oct 14, 2000 at 07:03:08PM -0400, Michael H. Warfield wrote:
=>> > >
=>> > > Note... Right now, dhcpcd still seems to have one feature that
=>> > > I seriously need and pump doesn't seem to have. It dumps all its lease
=>> > > information into a known file (/etc/dhcpc/dhcpcd-${IFACE}.info) which
=>> > > can be parsed by other scripts and compared with previous values. It's
=>> > > even in a convention form that can be "sourced" from a script to set
=>> > > environment variables. If the newer versions of pump support something
=>> > > similar, I haven't found it yet.
=>> >
=>> > You could conceivably set a script in pump.conf which does a
=>> > pump -s >> $FILE:
=>> >
=>> > Device eth0
=>> > IP: 216.78.197.8
=>> > Netmask: 255.255.252.0
=>> > Broadcast: 216.78.199.255
=>> > Network: 216.78.196.0
=>> > Boot server 205.152.133.254
=>> > Next server 0.0.0.0
=>> > Gateway: 216.78.196.1
=>> > Boot file: 005004a87713
=>> > Domain: sdf.bellsouth.net
=>> > Nameservers: 205.152.133.254 205.152.0.5
=>> > Renewal time: Sun Oct 15 05:42:22 2000
=>> > Expiration time: Sun Oct 15 07:12:22 2000
=>> >
=>> > Is that close? Just trying this with a forced renewal (pump -R), and
=>> > the script does not get run at all :(
=>> >
=>> > --
=>> > Hal B
=>> > [EMAIL PROTECTED]
=>> > [EMAIL PROTECTED]
=>> > [EMAIL PROTECTED]
=>
=> Mike
=>
_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list