FreeBSD: Skip inet6 addresses matching ::1 #1838
Conversation
|
@feld please add DCO |
| @@ -68,6 +68,7 @@ | |||
| end | |||
| end | |||
| if line =~ /\s+inet6 ([a-f0-9\:]+)%?(\w*)\s+prefixlen\s+(\d+)\s*\w*\s*([\da-fx]*)/ | |||
| next if $1 == "::1" | |||
There was a problem hiding this comment.
I'm not sure if this is the right approach. Is ::1 the only inet6 line in your ifconfig?
There was a problem hiding this comment.
No, I have fe80 addresses on all the interfaces and those are properly excluded.
Sometimes ::1 is on lo1 and not lo0, but it should always see my inet6 addresses starting with 2600 first and it doesn't pick them up. It always takes the ::1 as the default address for some strange reason.
On a very vanilla out-of-the-box FreeBSD install with IPv6 enabled Chef will pick up ::1 as the IPv6 address. This is the only way I found to fix it.
tpowell-progress
added
the
Status: Waiting on Contributor
feld
force-pushed
the
freebsd-ipv6-addr
branch
from
9ea4383 to
78e7f34
Compare
Signed-off-by: Mark Felder <[email protected]>
feld
force-pushed
the
freebsd-ipv6-addr
branch
from
78e7f34 to
cfadfcb
Compare
|
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
|
After writing #1841 I THINK I know what the problem is here. FreeBSD reports no scope for global addresses. Further, it uses scopeIDs, not scopes, so the sorting that happens here doesn't work at all. So the right fix here is to modify this code to:
|
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
This adds unittests for the FreeBSD network plugin. It also fixes various issues: * It now running on modern freebsd which has extra columns in netstat output. * It now properly reports encapuslation, MTU, and metric It also includes a test case with `::1` on `lo1` to start building out support for sorting out chef#1838 Signed-off-by: Phil Dibowitz <[email protected]>
|
@feld - ping? |
|
@feld following up |
|
@dafyddcrosby wants to look at this - will wait to close. |
If If I haven't tested this by December 3rd's community PR review (or if @feld responds), can close then |
@jaymzh Didn't get time over the weekend, and don't know when I will in the next few weeks, so let's close for now. |
|
Closing due to no response from original author. |
Fixes #1835