Adding helpful context to the response
Information Headers
Preliminaries
This note describes the information-headers module and how it can provide contextual information to make standard HTTP status codes more meaningful.
Every HTTP response has a three-digit status code. Their meanings are defined in various IETF RFCs. Oftentimes a response code is too generic and provides insufficient context to know how to troubleshoot a problem. The server's information-headers modules can provide that context.
An information header is a HTTP response header that begins with "rw-". The value associated with the header contains two parts: V1, an actual value; and V2, an expected value.
For example, if the rw-no-acceptable-type header contains a value like "V1=text/xml; V2=text/html, application/javascript", it means that the browser asked for a resource with the MIME-type shown in the first variable, V1 "text/xml", but the server is configured to only provide resources with the MIME-types shown in the second variable, V2 "text/html, application/javascript".
Some information headers only contain a single value when that is more germane; and some headers are self-explanatory and completely omit any value.
Refer to the other notes in this series to learn what conditions trigger each of these information headers.
rw-bad-range |
rw-blue-cache-add |
rw-blue-cache-dependency |
rw-blue-cache-hit |
rw-blue-cache-stale |
rw-blue-dir-deleted |
rw-blue-file-deleted |
rw-blue-invalid-option |
rw-blue-dep-deleted |
rw-blue-res-deleted |
rw-blue-lnk-deleted |
rw-blue-unsupported-method |
rw-conditional-range-failed |
rw-content-decoding |
rw-content-decoding |
rw-content-length-limit |
rw-content-length-missing |
rw-content-range |
rw-content-type-mismatch |
rw-content-type-missing |
rw-content-type-not-supported |
rw-counters-unknown-extension |
rw-counters-internal-error |
rw-encoding-cache-add |
rw-encoding-cache-add |
rw-encoding-cache-hit |
rw-encoding-dir-deleted |
rw-encoding-file-deleted |
rw-forbidden |
rw-header-not-allowed |
rw-hostname |
rw-if-match |
rw-if-none-match |
rw-if-unmodified-since |
rw-ip-access |
rw-language-not-acceptable |
rw-language-not-configured |
rw-method-not-allowed |
rw-no-acceptable-type |
rw-no-cache-control |
rw-origin-not-allowed |
rw-path |
rw-range-units-not-supported |
rw-rbac-expired |
rw-rbac-forbidden |
rw-rbac-forged |
rw-rbac-internal-error |
rw-rbac-missing-credentials |
rw-rbac-no-matching-role |
rw-rbac-no-resource-rule |
rw-rbac-remote-address |
rw-rbac-renewal |
rw-rbac-unsupported-action |
rw-rbac-unsupported-content-type |
rw-rbac-unsupported-method |
rw-rbac-unsupported-scheme |
rw-referer-nopush |
rw-request-dir-deleted |
rw-request-file-deleted |
rw-request-file-not-deleted |
rw-resource-path-illegal |
rw-resource-path-length |
rw-router-missing-module |
rw-user-agent-nopush |
rw-user-agent-noserve |
Configuration
Like all of the server's extra modules, the information-headers module is not enabled unless explicitly set to "on" in the modules section.
EBNF
| SP | ::= | U+20 |
| CR | ::= | U+0D |
| LEFT-CURLY-BRACKET | ::= | U+7B |
| RIGHT-CURLY-BRACKET | ::= | U+7D |
| module-name | ::= | 'information-headers' |
| named-module-entry | ::= | module-name SP ('on' | 'off') CR |
| modules-section | ::= | 'modules' SP LEFT-CURLY-BRACKET CR named-module-entry* RIGHT-CURLY-BRACKET CR |
Cookbook
Example 1: Information headers enabled
server {
modules {
information-headers on
}
}
Review
Key points to remember:
- It is recommended to keep this module turned on both during development and while in production.
- There is no performance penalty and no known security issues with having this module on.
