Community Documentation
HTTP Global Client settings
Document updated on Jan 18, 2022
When KrakenD communicates using HTTP with all your upstream services, it implements a concurrent-safe round tripper that supports HTTP, HTTPS, and HTTP proxies, and it caches connections for future re-use. This may leave many open connections when accessing many hosts. You can change the behavior of the transport layer using several settings presented below.
The following settings affect all connections from KrakenD to your services.
If you want to customize any of the settings below, they must be written at the top level of the configuration.
Fields of "false"
| Allows to set specific transport settings when using TLS in your upstream services. See TLS Client for more settings |
| Specifies the length of time to wait before spawning a RFC 6555 Fast Fallback connection. If zero, a default delay of 300ms is used. Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "300ms" |
| The interval between keep-alive probes for an active network connection. If zero, keep-alive probes are sent with a default value (currently 15 seconds), if supported by the protocol and operating system. Network protocols or operating systems that do not support keep-alives ignore this field. If negative, keep-alive probes are disabled. Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "15s" |
| The timeout of the dial function for creating connections.The default is no timeout. With or without a timeout, the operating system may impose its own earlier timeout. Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "0s" |
| When true prevents requesting compression with an Accept-Encoding: gzip request header when the Request contains no existing Accept-Encoding value. If the Transport requests gzip on its own and gets a gzipped response, it’s transparently decoded. However, if the user explicitly requested gzip it is not automatically uncompressed.Defaults to false |
| When true it disables HTTP keep-alives and will only use the connection to the server for a single HTTP request. Defaults to false |
| Only RESTful URL patterns are valid to access backends. Set to true if your backends aren’t RESTful, e.g.: /url.{some_variable}.jsonDefaults to false |
| If non-zero, specifies the amount of time to wait for a server’s first response headers after fully writing the request headers if the request has an Expect: 100-continue header. Zero means no timeout and causes the body to be sent immediately, without waiting for the server to approve. This time does not include the time to send the request header.Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "0s" |
| The maximum number of idle (keep-alive) connections across all hosts. Zero means no limit. Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "0s" |
| The maximum number of idle (keep-alive) connections across all hosts. Zero means no limit. Defaults to 0 |
| If non-zero, controls the maximum idle (keep-alive) connections to keep per-host. If zero, 250 is used instead.Defaults to 250 |
| If non-zero, specifies the amount of time to wait for a server’s response headers after fully writing the request (including its body, if any). This time does not include the time to read the response body. Specify units using ns (nanoseconds), us or µs (microseconds), ms (milliseconds), s (seconds), m (minutes), or h (hours).Defaults to "0s" |
Finally, the TLS Handshake Timeout is hardcoded to 10 seconds and cannot be changed.
Override settings using environment vars
When you declare in the configuration file any of the HTTP server or transport settings declared above, you can override its value through environment variables when starting the server.
All the environment variables have the same name as the settings above in uppercase and with the KRAKEND_ prefix. The following env vars are available:
KRAKEND_DIALER_TIMEOUTKRAKEND_DIALER_KEEP_ALIVEKRAKEND_DIALER_FALLBACK_DELAYKRAKEND_DISABLE_COMPRESSIONKRAKEND_DISABLE_KEEP_ALIVESKRAKEND_MAX_IDLE_CONNECTIONSKRAKEND_MAX_IDLE_CONNECTIONS_PER_HOSTKRAKEND_IDLE_CONNECTION_TIMEOUTKRAKEND_RESPONSE_HEADER_TIMEOUTKRAKEND_EXPECT_CONTINUE_TIMEOUT
You can start KrakenD with the desired variables to override what you have in the configuration:
$KRAKEND_MAX_IDLE_CONNECTIONS_PER_HOST=200 krakend run -c krakend.jsonMax IDLE connections
Having a high number of IDLE connections to every backend affects directly to the performance of the proxy layer. This is why you can control the number using the max_idle_connections setting. For instance:
{
"version": 3,
"max_idle_connections": 150
}
KrakenD will close connections sitting idle in a “keep-alive” state when max_idle_connections is reached. If no value is set in the configuration file, KrakenD will use 250 by default.
Every ecosystem needs its own setting, have this in mind:
- If you set a number very high for
max_idle_connectionsyou might exhaust your system’s port limit. - If you set a number very low, new connections will be frequently created and a low rate of connection reuse will take place.
