Source: autofs Version: 5.1.8-3 Severity: important Tags: upstream X-Debbugs-Cc: car...@debian.org Control: found -1 5.1.8-2+deb12u1 Control: found -1 5.1.8-2 Control: found -1 5.1.7-1+deb11u1 Control: found -1 5.1.7-1 Control: tags -1 + bullseye bookworm trixie sid Control: forwarded -1 https://www.spinics.net/lists/autofs/msg02668.html
Hi [ This was originally already handled upstream but filling as well a downstream bugreport for earsier handling for the respective bullseye-pu and bookworm-pu updates. ] After changes upstream for mounts from IPv6 addresses there were delays seen under certain circumstances. In the particular known case the server was dual stacked IPv4 and IPv6, the client in a IPv6 capable different subnet, equipped only with IPv4, but still having only for local-link address setup (address, default route on the interface). The IPv6 server address is not reachable and in this case the IPv6 address of the server was still probed when calculating the proximity. Upstream original report which needed some extra rounds for clarifications: https://www.spinics.net/lists/autofs/msg02643.html Upstream patch series for review: https://www.spinics.net/lists/autofs/msg02668.html After fixing in unstable, I'm willing to help the fix go as well in bookworm and bullseye via next point releases (possibly not anymore timely for the first bookworm point release). Regards, Salvatore
