Comment 4 for bug 1668771
Revision history for this message
| Drew Freiberger (afreiberger) wrote Re: systemd-resolved negative caching for extended period of time | #4 |
© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
73416ba
(Get the code!)
This affects bionic openstack cloud environments when os-*-hostname is configured for keystone, and the keystone entry is deleted temporarily from upstream dns, or the upstream dns fails providing no record for the lookup of keystone.
We have to then flush all caches across the cloud once DNS issue is resolved, rather than auto-healing at 60 seconds as if we were running nscd with negative-ttl set to 60 seconds.
Ultimately, a negative TTL that is settable would be ideal, or the ability to not cache negative hits would also be useful. Only workaround now is to not use caches or to operationally flush caches as needed.