middleware/cache/README.md

# cache

`cache` enables a frontend cache.

## Syntax

~~~
cache [ttl] [zones...]
~~~

* `ttl` max TTL in seconds. If not specified, the TTL of the reply (SOA minimum or minimum TTL in the
  answer section) will be used.
* `zones` zones it should cache for. If empty, the zones from the configuration block are used.

Each element in the cache is cached according to its TTL. For the negative cache, the SOA's MinTTL
value is used.

A cache mostly makes sense with a middleware that is potentially slow (e.g., a proxy that retrieves an
answer), or to minimize backend queries for middleware like etcd. Using a cache with the file
middleware essentially doubles the memory load with no conceivable increase of query speed.

The minimum TTL allowed on resource records is 5 seconds.

If monitoring is enabled (via the `prometheus` directive) then the following extra metrics are added:
* coredns_cache_hit_count_total, and
* coredns_cache_miss_count_total

They both work on a per-zone basis and just count the hit and miss counts for each query.

## Examples

~~~
cache 10
~~~

Enable caching for all zones, but cap everything to a TTL of 10 seconds.

~~~
proxy . 8.8.8.8:53
cache example.org
~~~

Proxy to Google Public DNS and only cache responses for example.org (or below).
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00			`# cache`

			`cache` enables a frontend cache.

			`## Syntax`

			`~~~`
			`cache [ttl] [zones...]`
			`~~~`

Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			* `ttl` max TTL in seconds. If not specified, the TTL of the reply (SOA minimum or minimum TTL in the
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00			`answer section) will be used.`
Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			* `zones` zones it should cache for. If empty, the zones from the configuration block are used.
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00
Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			`Each element in the cache is cached according to its TTL. For the negative cache, the SOA's MinTTL`
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00			`value is used.`

Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			`A cache mostly makes sense with a middleware that is potentially slow (e.g., a proxy that retrieves an`
			`answer), or to minimize backend queries for middleware like etcd. Using a cache with the file`
Doc updates Remove miek.nl in favor of example.org and other smaller cleanups. 2016-08-22 07:47:03 +01:00			`middleware essentially doubles the memory load with no conceivable increase of query speed.`
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00
Updates to README 2016-04-19 10:26:29 +00:00			`The minimum TTL allowed on resource records is 5 seconds.`

Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			If monitoring is enabled (via the `prometheus` directive) then the following extra metrics are added:
Add monitoring metrics docs to middleware/cache 2016-06-23 12:57:06 +00:00			`* coredns_cache_hit_count_total, and`
			`* coredns_cache_miss_count_total`

Update README.md More grammatical fixes. 2016-08-22 13:51:54 -07:00			`They both work on a per-zone basis and just count the hit and miss counts for each query.`
Add monitoring metrics docs to middleware/cache 2016-06-23 12:57:06 +00:00
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00			`## Examples`

			`~~~`
Updates to README 2016-04-19 10:26:29 +00:00			`cache 10`
			`~~~`

			`Enable caching for all zones, but cap everything to a TTL of 10 seconds.`

			`~~~`
			`proxy . 8.8.8.8:53`
Doc updates Remove miek.nl in favor of example.org and other smaller cleanups. 2016-08-22 07:47:03 +01:00			`cache example.org`
Cache (#126) * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compress and scrubbed so it will fit the specific client asking for it. * first simple test, less exporting of stuff * more * Add middleware/cache Add a caching middleware that caches nxdomain, nodata and successful responses. It differentiates between DNSSEC on normal DNS replies. Each reply is compressed and scrubbed so it will fit the specific client asking for it. The TTL is decremented with the time spend in the cache. There is syntax that allows you to cap the TTL for all records, no matter what. This allows for a shortlived cache, just to absorb query peaks. +Tests * cache test infrastructure * Testing 2016-04-19 11:13:24 +01:00			`~~~`

Doc updates Remove miek.nl in favor of example.org and other smaller cleanups. 2016-08-22 07:47:03 +01:00			`Proxy to Google Public DNS and only cache responses for example.org (or below).`