On 2026-09-23 12:00:10 +0200, Marco Moock wrote: > Am 23.09.26 um 11:50 schrieb Vincent Lefevre: > > On 2026-09-22 15:33:05 +0200, Marco Moock wrote: > > > Am 22.09.26 um 05:30 schrieb Max Nikulin: > > > > The issue is that nscd caches for an hour results with some IPv6 > > > > addresses and no IPv4 ones due to SERVFAIL in response to the A query. > > > > As a result various tools can not connect the host due to lack of global > > > > IPv6 routing. > > > > > > Not an IPv6 issue and not an nscd issue. > > > If nscd caches the positive (e.g. answer with A/AAAA record) or negative > > > (server replies that the requested record does not exist) answer, it is > > > operation correctly. The failing DNS servers need to be handled at > > > libnss_dns. > > > > Or perhaps nscd just misuses libnss_dns. Without documentation of > > how nscd uses this library exactly, who knows? > > The libraries are queried in order listed in nsswitch.conf. This can be > tuned (continue after a successful query, retry options etc.) with some > settings, see the nsswitch.conf manpage.
I repeat: "hosts: files dns". But given the fact that these are arbitrary remote hosts, "files" will not give an answer. So this yields "dns". But I don't know how this is handled. > nscd is just a cache between the requesting application and the actual > resolver libraries. You're repeating that, but tests show that it is not true. The IPv4 address is cached, but nscd does not always give it in its response with AF_UNSPEC. -- Vincent Lefèvre <[email protected]> - Web: <https://www.vinc17.net/> 100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/> Work: CR INRIA - computer arithmetic / Pascaline project (LIP, ENS-Lyon)
