diff options
| author |  Miek Gieben <miek@miek.nl>
| 2017-09-20 11:36:41 +0100 |
|---|---|---|
| committer |  GitHub <noreply@github.com>
| 2017-09-20 11:36:41 +0100 |
| commit | cd5879f866641141a241706d0d0a0219e52b7e68 (patch) | |
| tree | e88a040834128512a4682aebae5aeafa456588f1 /plugin/cache/handler.go | |
| parent | be4770927048627b59a6ce80587d668d7c028518 (diff) | |
| download | coredns-cd5879f866641141a241706d0d0a0219e52b7e68.tar.gz coredns-cd5879f866641141a241706d0d0a0219e52b7e68.tar.zst coredns-cd5879f866641141a241706d0d0a0219e52b7e68.zip | |
plugin/cache: cap TTL on first answer (#1092)
Cache would let the first response through and would then cap subsequent
ones to whatever the cache duration was. This would lead to huge drops
in TTL values: 3600 -> 20 for instance, which is not only bad, but can
mess up your careful TTL planning business.
This PR fixes that and applies the cache duration to all replies. As a
bonus I could remove a time.Sleep() from the cache test and just check
for the cache duration as the TTL on the reply.
Fixes #1038
Diffstat (limited to 'plugin/cache/handler.go')
| -rw-r--r-- | plugin/cache/handler.go | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/plugin/cache/handler.go b/plugin/cache/handler.go index ebd87d659..f3f2e675b 100644 --- a/plugin/cache/handler.go +++ b/plugin/cache/handler.go @@ -19,7 +19,7 @@ func (c *Cache) ServeDNS(ctx context.Context, w dns.ResponseWriter, r *dns.Msg) qtype := state.QType() zone := plugin.Zones(c.Zones).Matches(qname) if zone == "" { - return c.Next.ServeDNS(ctx, w, r) + return plugin.NextOrFailure(c.Name(), c.Next, ctx, w, r) } do := state.Do() // TODO(): might need more from OPT record? Like the actual bufsize? |