dhcpcd-discuss

Changed the test script to remove /var/db/dhcp* before test. Same results:

# cat dhcpcd.*
debug
logfile /xxx/dhcpcd.log
interface eth0
ia_na

Mar 29 05:45:56 dhcpcd-6.10.1 starting
Mar 29 05:45:56 forking to background
Mar 29 05:45:56 forked to background, child pid 515
Mar 29 05:45:56 eth0: executing `/libexec/dhcpcd-run-hooks' PREINIT
Mar 29 05:45:56 script_runreason: /libexec/dhcpcd-run-hooks: WEXITSTATUS 127
Mar 29 05:45:56 eth0: executing `/libexec/dhcpcd-run-hooks' CARRIER
Mar 29 05:45:56 script_runreason: /libexec/dhcpcd-run-hooks: WEXITSTATUS 127
Mar 29 05:45:56 DUID 00:01:00:01:1e:85:14:af:6c:a8:49:86:5c:59
Mar 29 05:45:56 eth0: IAID 49:86:5c:59
Mar 29 05:45:56 eth0: delaying IPv6 router solicitation for 0.9 seconds
Mar 29 05:45:56 eth0: reading lease `/var/db/dhcpcd-eth0.lease6'
Mar 29 05:45:56 eth0: soliciting a DHCPv6 lease
Mar 29 05:45:56 eth0: delaying SOLICIT6 (xid 0x4ead5b), next in 0.9 seconds
Mar 29 05:45:57 eth0: broadcasting SOLICIT6 (xid 0x4ead5b), next in 1.1
seconds
Mar 29 05:45:57 eth0: soliciting an IPv6 router
Mar 29 05:45:57 eth0: sending Router Solicitation
Mar 29 05:45:58 eth0: broadcasting SOLICIT6 (xid 0x4ead5b), next in 2.1
seconds
Mar 29 05:46:00 eth0: broadcasting SOLICIT6 (xid 0x4ead5b), next in 4.1
seconds
Mar 29 05:46:01 eth0: sending Router Solicitation
Mar 29 05:46:04 eth0: broadcasting SOLICIT6 (xid 0x4ead5b), next in 8.0
seconds
Mar 29 05:46:05 eth0: sending Router Solicitation
Mar 29 05:46:06 received SIGTERM, stopping
Mar 29 05:46:06 eth0: removing interface
Mar 29 05:46:06 eth0: executing `/libexec/dhcpcd-run-hooks' STOPPED
Mar 29 05:46:06 script_runreason: /libexec/dhcpcd-run-hooks: WEXITSTATUS 127
Mar 29 05:46:06 dhcpcd exited
#


On Tue, Mar 29, 2016 at 12:59 PM, Roy Marples <roy@xxxxxxxxxxxx> wrote:

> On 29/03/2016 17:40, Shahid Mahmood wrote:
> > Certainly; TAHI test script deletes old leases prior to starting a new
> > test. I can remove that from the test script itself, but not sure what
> > else it will do to the rest of the tests. I guess only one way to find
> out.
>
> No no.
> I suspect the test needs to delete the old lease for it to work - maybe
> it forces the server not to confirm any prior lease, I don't know.
> I suggest either recompiling dhcpcd to match TAHI expectations, or
> adjust the TAHI script to delete from same directory dhcpcd actually uses.
>
> Also, if there is a test that wants to look at files setup for say DNS
> then you'll need to correct the script path in the same way.
>
> Roy
>