This PR adds the *root* middleware that specifies a path where
all zone file (the *file* middleware is the only consumer now) can
be found. It works the same as in Caddy.
Documentation can be found in the README.md of the middleware.
Fixes#307
Make DefaultPort just be 53. And use the value of Port to make the zone
stanzas complete (instead of defaulting to 53 always). This allows you
to override the port with dns.port.
For some reasons there was a dnsserver/middleware.go that defined
the middleware handlers. This code was a repeat from
middleware/middleware.go. Removed dnsserver/middleware.go and replaced
all uses of dnsserver.Middleware with middleware.Middleware.
Added dnsserver/address_test.go to test the zone normalization (and to
improve the test coverage). The deleted file will also improve the test
coverage :)
When no Corefile is given, default to loading the whoami middleware on
the default port (2053). Also add back the -port flag that allows you
to override the default port.
Further cleanup the startup messages and use caddy's OnStartupComplete()
to blurp out which zones and ports we have. These can be suppressed
with the -quiet flag.
Normal startup:
miek.nl.:1053
miek.nl2.:1053
example.org.:1054
2016/09/17 20:41:19 [INFO] CoreDNS-001 starting
CoreDNS-001 starting
with the -quiet flag:
2016/09/17 20:41:34 [INFO] CoreDNS-001 starting
Add a new middleware that tells you who you are; IP, port and transport
is echoed back.
Also some various cleanup and documentation improvements while at it:
* ResponseWriter: improve the documentation of these helper functions.
* And add an NextHandler for use in tests. Make chaos_test.go and
* whoam_test.go use it.
Stop the caddy message and start our own init notifications.
Log the version of CoreDNS when starting up.
Fix all middleware's setup functions so that return the error prefixed
with *which* middleware was failing; leads to better debuggable errors
when starting up.
Move all (almost all) Go files in middleware into their
own packages. This makes for better naming and discoverability.
Lot of changes elsewhere to make this change.
The middleware.State was renamed to request.Request which is better,
but still does not cover all use-cases. It was also moved out middleware
because it is used by `dnsserver` as well.
A pkg/dnsutil packages was added for shared, handy, dns util functions.
All normalize functions are now put in normalize.go
See
17709a7d3f
where there was a slight change (for the better) on how to register.
Fix the CoreDNS to adhire to this. Needs Caddy from master to compile at
this moment.
* Set version to 001
* Remove k8stest, test fails is k8s is not there: touch luck
* Remove server directory: not used anymore
* Disable k8s test (for now)
* gometalinter changes
* Make CoreDNS a server type plugin for Caddy
Remove code we don't need and port all middleware over. Fix all tests
and rework the documentation.
Also make `go generate` build a caddy binary which we then copy into
our directory. This means `go build`-builds remain working as-is.
And new etc instances in each etcd test for better isolation.
Fix more tests and rework test.Server with the newer support Caddy offers.
Fix Makefile to support new mode of operation.
* Adding parsing for label selector to Corefile
* Updating comment typo in k8sCorefile
* Adding implementation of label support to filter exposed objects
* Updating TODO list
Make the receiver a pointer so that the uptdateStubZones map update will
retain the stubzones found, unlike the current case where the update
will be applied and then promptly forgotten, because it is working on a
copy.
Add test/etcd_test.go to test a large part of the code. This didn't
catch the chaos middleware hack though. The chaos middleware zones are
now *not* automatically added. You have to take care of that by yourself
(docs updates).
When using debug queries and falling through to the next middleware in
etcd, restore the original (with o-o.debug) query before passing it on.
* Removing old unused inline k8s API code and tests.
* Adding parsing implementation for `resyncperiod` keyword from Corefile.
* Adding tests for parsing `resyncperiod` keyword from Corefile.
8 Updating README.md and conf/k8sCorefile.
* Merge notification code by @aledbf and update for recent changes.
* Fix travis environment to correctly build with k8s.io and forked repositories.
* Refactored kubernetes Corefile parser
* Added lots of Corefile parsing tests
* Updating comment to remove references to Caddy
* Updating README to mark TODO items complete
* Changing all debug print statements over to use logging package
* Commenting out unused functions. TODO: remove when it is not needed
* Update README with namespace and template example
* Adding note about changing the record name format via a template
* Adding test scripts to automate k8s startup
* Automating k8s namespace creation
* Adding automation to start 4 k8s services
* Updating documentation for k8s tests
* Avoid downloading kubectl if already exists
* Adding debug statement when namespace is not exposed.
* Adding basic kubernetes integration tests
* Makefile now contains a "testk8s" target. This target requires k8s to
be running.
* Adding test/kubernetes_test.go file with a couple of basic A record
tests.
* Updating k8s integration tests to only run k8s integration tests
* Adding support for namespace wildcards
* Refactoring to move filtering logic to kubernetes.go file
* go fmt fixes
* Adding wildcard support for namespaces and service names
* Kubernetes integration tests updated for A records.
* Expanded record name assembly for answer section not yet implemented.
* Refactoring to focus k8sclient code just on accessing k8s API.
Filtering now handled in kubernetes.go
* Adding wildcard test cases
* Adding skydns startup script. (To allow side by side testing of wildcards.)
* Commenting out record name assmebly based on NameTemplate. Need to improve template before this makes sense.
* Adding basic SRV integration tests
* Need to add verification for additional answer section
* Fixing comments and formatting
* Moving wildcard constants to vars
* Travis test execution appears to be failing on access to these
constants
* Fixing access to util package
* Trying to work around Travis test bug
* Reverting to access kubernetes/util as "util"
Travis breakage is due to "Infoblox-CTO" in src path
* Removing unnecessary gitignore pattern
* Updating Makefile to run unittests for subpackages
* Adding Corefile validation to ignore overlapping zones
* Fixing SRV query handling
* Updating README.md now that SRV works
* Fixing debug message, adding code comment
* Clarifying implementation of zone normalization
* "Overlapping zones" is ill-defined. Reimplemented zone overlap/subzone
checking to contain these functions in k8s middleware and provide
better code comments explaining the normalization.
* Separate build verbosity from test verbosity
* Cleaning up comments to match repo code style
* Merging warning messages into single message
* Moving function docs to before function declaration
* Adding test cases for k8sclient connector
* Tests cover connector create and setting base url
* Fixed bugs in connector create and setting base url functions
* Updaing README to group and order development work
* Priority focused on achieving functional parity with SkyDNS.
* Adding work items to README and cleaning up formatting
* More README format cleaning
* List formating
* Refactoring k8s API call to allow dependency injection
* Add test cases for data parsing from k8s into dataobject structures
* URL is dependency-injected to allow replacement with a mock http
server during test execution
* Adding more data validation for JSON parsing tests
* Adding test case for GetResourceList()
* Adding notes about SkyDNS embedded IP and port record names
* Marked test case implemented.
* Fixing formatting for example command.
* Fixing formatting
* Adding notes about Docker image building.
* Adding SkyDNS work item
* Updating TODO list
* Adding name template to Corefile to specify how k8s record names are assembled
* Adding template support for multi-segment zones
* Updating example CoreFile for k8s with template comment
* Misc whitespace cleanup
* Adding SkyDNS naming notes
* Adding namespace filtering to CoreFile config
* Updating example k8sCoreFile to specify namespaces
* Removing unused codepath
* Adding check for valid namespace
* More README TODO restructuring to focus effort
* Adding template validation while parsing CoreFile
* Record name template is considered invalid if it contains a symbol of the form ${bar} where the symbol
"${bar}" is not an accepted template symbol.
* Refactoring generation of answer records
* Parse typeName out of query string
* Refactor answer record creation as operation over list of ServiceItems
* Moving k8s API caching into SkyDNS equivalency segment
* Adding function to assemble record names from template
* Warning: This commit may be broken. Syncing to get laptop code over to dev machine.
* More todo notes
* Adding comment describing sample test data.
* Update k8sCorefile
* Adding comment
* Adding filtering support for kubernetes "type"
* Required refactoring to support reuse of the StringInSlice function.
* Cleaning up formatting
* Adding note about SkyDNS supporting word "any".
* baseUrl -> baseURL
* Also removed debug statement from core/setup/kubernetes.go
* Fixing test breaking from Url -> URL naming changes
* Changing record name template language ${...} -> {...}
* Fix formatting with go fmt
* Updating all k8sclient data getters to return error value
* Adding error message to k8sclient data accessors
* Cleaning up setup for kubernetes
* Removed verbose nils in initial k8s middleware instance
* Set reasonable defaults if CoreFile has no parameters in the
kubernetes block. (k8s endpoint, and name template)
* Formatting cleanup -- go fmt
* Laying down kubernetes middleware foundation
* Duplicated a bunch of code form etcd middleware
* Duplicated code hacked to compile and load as a separate middleware
* Adding verbose build option to Makefile
* Removing stubzone and tls support
tls and stubzone support was carried over from base etcd middleware code.
Removing to simplify the kube middleware implementation. (For now.)
* Adding conf directory for sample conf files
* Removing stubzone support from query handler
* Remove upstream and proxy from k8s corefile.
Not sure that upstream or proxy makes sense for a k8s backed zone.
* Comment out use of singleflight serialization
* Removing parsing support for "upstream" directive from k8s
* Removing upstream directive parsing code
* Removing CNAME and TXT lookup implementation
* Create README.md
Brain-dump of DNS record name assembly and open work items.
* Adding notes about wildcard handling
* Adding basic k8s API client
* Fleshing out methods on k8s connector
* Remove PathPrefix from middleware init
* Removing incorrect plural
* Adding brute-force k8s service lookup functions
* Initializing k8s API connector during startup
* Hacking around to call k8s connector
* Parsing incoming domain name into serviceName and namespace
* Improving and simplifying k8s zone matching and label segmentation
* Removing unused functions carried over from etcd middleware
* Adding basic return of k8s data to DNS client
* updated debugging println statements to flag with "[debug]"
* removed code in kubernetes.go::Records that was a hold-over from etcd middleware.
* Removed some random exploratory hacking.
* Minior README.md updates
* Updating with demo instructions
* Updating README.md with CoreFile and removing completed TODO items
* Updating conf file and README to reflect DNS response cache works
* Disabling DNS response caching
* Adding debug statement on entry to Records()
* Changing port number in exampes to port 53.
* Misc style and clarity changes
* Removing empty function definitions
* Adding comment to track future cleanup
* Refactoring README to follow style of other middleware
* Exposing dataobject field (typo)
With this you can retreive the raw data that the etcd middleware
used to create the reply. The debug data is put in TXT records
that are stuffed in the CH classs. This is only enabled if you
specify `debug` in the etcd stanza.
You can retrieve it by prefixing your query with 'o-o.debug.'
For instance:
; <<>> DiG 9.10.3-P4-Ubuntu <<>> @localhost -p 1053 SRV o-o.debug.production.*.skydns.local
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 47798
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 3
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;o-o.debug.production.*.skydns.local. IN SRV
;; ANSWER SECTION:
production.*.skydns.local. 154 IN SRV 10 50 8080 service1.example.com.
production.*.skydns.local. 154 IN SRV 10 50 8080 service2.example.com.
;; ADDITIONAL SECTION:
skydns.local.skydns.east.production.rails.1. 154 CH TXT "service1.example.com:8080(10,0,,false)[0,]"
skydns.local.skydns.west.production.rails.2. 154 CH TXT "service2.example.com:8080(10,0,,false)[0,]"
Make middleware that sets up a (http) handler survive a graceful
restart. We calls the middleware's Shutdown function(s). If restart
fails the Start function is called again.
* middleware/health: OK
* middleware/pprof: OK
* middleware/metrics: OK
All restart OK.
Fix CoreDNS graceful reloading. This uses the same stuff as Caddy
(obviously), but extends it for UDP listeners as well. Also add to the
README that we *will* call Shutdown for middleware.
Fixes#4
* Changed reference to Caddy over to CoreDNS
* Removing references to caddy
* Fixed misleading error message to reference coredns
* Cleaning up references to caddy
* Adding clean and deps targets
Muscle memory is resulting in "make clean" commands.
* Adding test target to makefile
* More "Caddy" cleanup
Put error back in the correct place in the directives.go. Also don't
make it a pointer. If it *is* a pointer the buildstack function does
not correctly set the Next Handler. Don't understand *why* this is
different from Caddy. Anyway this fixes it, with the caveat that
the error log file is now openend earlier in the startup.
Fixes#127
* 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
