A test-server filter which is capable of generating test responses.
bazel test -c dbg //test/server:http_test_server_filter_integration_test
bazel build -c opt :nighthawk_test_server
It is possible to enable additional envoy extension by adding them here before the build. By default, Nighthawk's test server is set up with the minimum extension set needed for it to operate as documented.
test-server.yaml
sample content
static_resources:
listeners:
# define an origin server on :10000 that always returns "lorem ipsum..."
- address:
socket_address:
address: 0.0.0.0
port_value: 10000
filter_chains:
- filters:
- name: envoy.filters.network.http_connection_manager
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
generate_request_id: false
codec_type: AUTO
stat_prefix: ingress_http
route_config:
name: local_route
virtual_hosts:
- name: service
domains:
- "*"
http_filters:
- name: dynamic-delay
typed_config:
"@type": type.googleapis.com/nighthawk.server.DynamicDelayConfiguration
static_delay: 0.5s
- name: test-server # before envoy.router because order matters!
typed_config:
"@type": type.googleapis.com/nighthawk.server.ResponseOptions
response_body_size: 10
v3_response_headers:
- { header: { key: "foo", value: "bar" } }
- {
header: { key: "foo", value: "bar2" },
append: true,
}
- { header: { key: "x-nh", value: "1" } }
- name: envoy.filters.http.router
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
dynamic_stats: false
admin:
access_log_path: /tmp/envoy.log
address:
socket_address:
address: 0.0.0.0
port_value: 8081
The ResponseOptions proto used to be
shared by the Test Server
, Dynamic Delay
, and Time Tracking
filter
extensions. Now, each of these filters has its own primary configuration, respectively:
- ResponseOptions for
Test Server
- DynamicDelayConfiguration for
Dynamic Delay
- TimeTrackingConfiguration for
Time Tracking
However, currently, each filter still uses the same x-nighthawk-test-server-config
request-header, which is a ResponseOptions proto. When this header is provided, each
filter will interpret only the parts that are relevant to it.
response_body_size
- number of 'a' characters repeated in the response body.response_headers
- list of headers to add to response. Ifappend
is set totrue
, then the header is appended.echo_request_headers
- if set totrue
, then append the dump of request headers to the response body.
The response options above could be used to test and debug proxy or server configuration, for example, to verify request headers that are added by intermediate proxy:
$ curl -6 -v [::1]:8080/nighthawk
* Trying ::1:8080...
* TCP_NODELAY set
* Connected to ::1 (::1) port 8080 (#0)
> GET /nighthawk
> Host: [::1]:8080
> User-Agent: curl/7.68.0
> Accept: */*
>
* Mark bundle as not supporting multiuse
< HTTP/1.1 200 OK
< content-length: 254
< content-type: text/plain
< foo: bar
< foo: bar2
< x-nh: 1
< date: Wed, 03 Jun 2020 13:34:41 GMT
< server: envoy
< x-service: nighthawk_cluster
< via: 1.1 envoy
<
aaaaaaaaaa
Request Headers:
':authority', '[::1]:8080'
':path', '/nighthawk'
':method', 'GET'
':scheme', 'https'
'user-agent', 'curl/7.68.0'
'accept', '*/*'
'x-forwarded-proto', 'http'
'via', '1.1 google'
'x-forwarded-for', '::1,::1'
* Connection #0 to host ::1 left intact
This example shows that intermediate proxy has added x-forwarded-proto
and
x-forwarded-for
request headers.
The Dynamic Delay interprets the DynamicDelayConfiguration
for startup configuration, and the
oneof_delay_options
part in the ResponseOptions proto for
x-nighthawk-test-server-config
. If specified, it can be used to:
- Configure a static delay via
static_delay
. - Configure a delay which linearly increase as the number of active requests grows, representing a simplified model of an overloaded server, via
concurrency_based_linear_delay
.
All delays have a millisecond-level granularity.
At the time of writing this, there is a known issue with merging configuration provided via request headers into the statically configured configuration. The current recommendation is to use either static, or dynamic configuration (delivered per request header), but not both at the same time.
# If you already have Envoy running, you might need to set --base-id to allow the test-server to start.
➜ /bazel-bin/nighthawk/source/server/server --config-path /path/to/test-server.yaml
# Verify the test server with a curl command similar to:
➜ curl -H "x-nighthawk-test-server-config: {response_body_size:20, static_delay: \"0s\"}" -vv 127.0.0.1:10000
➜ bazel-bin/nighthawk_test_server --help
USAGE:
bazel-bin/nighthawk_test_server [--stats-tag <string>] ...
[--enable-core-dump] [--socket-mode
<string>] [--socket-path <string>]
[--disable-extensions <string>]
[--cpuset-threads]
[--enable-mutex-tracing]
[--disable-hot-restart] [--mode
<string>] [--parent-shutdown-time-s
<uint32_t>] [--drain-strategy <string>]
[--drain-time-s <uint32_t>]
[--file-flush-interval-msec <uint32_t>]
[--service-zone <string>]
[--service-node <string>]
[--service-cluster <string>]
[--hot-restart-version]
[--restart-epoch <uint32_t>]
[--log-path <string>]
[--enable-fine-grain-logging]
[--log-format-escaped] [--log-format
<string>] [--component-log-level
<string>] [-l <string>]
[--local-address-ip-version <string>]
[--admin-address-path <string>]
[--ignore-unknown-dynamic-fields]
[--reject-unknown-dynamic-fields]
[--allow-unknown-static-fields]
[--allow-unknown-fields] [--config-yaml
<string>] [-c <string>] [--concurrency
<uint32_t>] [--base-id-path <string>]
[--skip-hot-restart-parent-stats]
[--skip-hot-restart-on-no-parent]
[--use-dynamic-base-id] [--base-id
<uint32_t>] [--] [--version] [-h]
Where:
--stats-tag <string> (accepted multiple times)
This flag provides a universal tag for all stats generated by Envoy.
The format is ``tag:value``. Only alphanumeric values are allowed for
tag names. For tag values all characters are permitted except for '.'
(dot). This flag can be repeated multiple times to set multiple
universal tags. Multiple values for the same tag name are not
allowed.
--enable-core-dump
Enable core dumps
--socket-mode <string>
Socket file permission
--socket-path <string>
Path to hot restart socket file
--disable-extensions <string>
Comma-separated list of extensions to disable
--cpuset-threads
Get the default # of worker threads from cpuset size
--enable-mutex-tracing
Enable mutex contention tracing functionality
--disable-hot-restart
Disable hot restart functionality
--mode <string>
One of 'serve' (default; validate configs and then serve traffic
normally) or 'validate' (validate configs and exit).
--parent-shutdown-time-s <uint32_t>
Hot restart parent shutdown time in seconds
--drain-strategy <string>
Hot restart drain sequence behaviour, one of 'gradual' (default) or
'immediate'.
--drain-time-s <uint32_t>
Hot restart and LDS removal drain time in seconds
--file-flush-interval-msec <uint32_t>
Interval for log flushing in msec
--service-zone <string>
Zone name
--service-node <string>
Node name
--service-cluster <string>
Cluster name
--hot-restart-version
Hot restart compatibility version
--restart-epoch <uint32_t>
Hot restart epoch #
--log-path <string>
Path to logfile
--enable-fine-grain-logging
Logger mode: enable file level log control (Fine-Grain Logger) or not
--log-format-escaped
Escape c-style escape sequences in the application logs
--log-format <string>
Log message format in spdlog syntax (see
https://github.com/gabime/spdlog/wiki/3.-Custom-formatting)
Default is "[%Y-%m-%d %T.%e][%t][%l][%n] [%g:%#] %v"
--component-log-level <string>
Comma-separated list of component log levels. For example
upstream:debug,config:trace
-l <string>, --log-level <string>
Log levels: [trace][debug][info][warning
|warn][error][critical][off]
Default is [info]
--local-address-ip-version <string>
The local IP address version (v4 or v6).
--admin-address-path <string>
Admin address path
--ignore-unknown-dynamic-fields
Ignore unknown fields in dynamic configuration
--reject-unknown-dynamic-fields
Reject unknown fields in dynamic configuration
--allow-unknown-static-fields
Allow unknown fields in static configuration
--allow-unknown-fields
Allow unknown fields in static configuration (DEPRECATED)
--config-yaml <string>
Inline YAML configuration, merges with the contents of --config-path
-c <string>, --config-path <string>
Path to configuration file
--concurrency <uint32_t>
# of worker threads to run
--base-id-path <string>
Path to which the base ID is written
--skip-hot-restart-parent-stats
When hot restarting, by default the child instance copies stats from
the parent instance periodically during the draining period. This can
potentially be an expensive operation; set this to true to reset all
stats in child process.
--skip-hot-restart-on-no-parent
When hot restarting with epoch>0, the default behavior is for the
child to crash if the connection to the parent cannot be established.
Set this to true to instead continue with a regular startup, while
retaining the new epoch value.
--use-dynamic-base-id
The server chooses a base ID dynamically. Supersedes a static base ID.
May not be used when the restart epoch is non-zero.
--base-id <uint32_t>
Base ID so that multiple envoys can run on the same host if needed
--, --ignore_rest
Ignores the rest of the labeled arguments following this flag.
--version
Displays version information and exits.
-h, --help
Displays usage information and exits.
envoy