homelab/traefik
1
0
Fork 0
Code Activity

Add health check timeout parameter

Browse source
This commit is contained in:
Jared Biel 2018-09-27 13:16:03 -05:00 • committed by Traefiker Bot
parent f10516deb7
commit 5acd43efaf
45 changed files with 189 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

12
docs/basics.md
View file

@ -452,9 +452,11 @@ If not, a new backend will be assigned.
#### Health Check
A health check can be configured in order to remove a backend from LB rotation as long as it keeps returning HTTP status codes other than `2xx` or `3xx` to HTTP GET requests periodically carried out by Traefik.
The check is defined by a path appended to the backend URL and an interval (given in a format understood by [time.ParseDuration](https://golang.org/pkg/time/#ParseDuration)) specifying how often the health check should be executed (the default being 30 seconds).
Each backend must respond to the health check within 5 seconds.
A health check can be configured in order to remove a backend from LB rotation as long as it keeps returning HTTP status codes other than `2xx` or `3xx` to HTTP GET requests periodically carried out by Traefik.
The check is defined by a path appended to the backend URL and an interval specifying how often the health check should be executed (the default being 30 seconds.)
Each backend must respond to the health check within a timeout duration (the default being 5 seconds.)
Interval and timeout are to be given in a format understood by [time.ParseDuration](https://golang.org/pkg/time/#ParseDuration).
The interval must be greater than the timeout. If configuration doesn't reflect this, the interval will be set to timeout + 1 second.
By default, the port of the backend server is used, however, this may be overridden.
A recovering backend returning `2xx` or `3xx` responses again is being returned to the LB rotation pool.
@ -466,6 +468,7 @@ For example:
[backends.backend1.healthcheck]
path = "/health"
interval = "10s"
timeout = "3s"
```
To use a different port for the health check:
@ -475,6 +478,7 @@ To use a different port for the health check:
[backends.backend1.healthcheck]
path = "/health"
interval = "10s"
timeout = "3s"
port = 8080
```
@ -486,6 +490,7 @@ To use a different scheme for the health check:
[backends.backend1.healthcheck]
path = "/health"
interval = "10s"
timeout = "3s"
scheme = "http"
```
@ -496,6 +501,7 @@ Additional http headers and hostname to health check request can be specified, f
[backends.backend1.healthcheck]
path = "/health"
interval = "10s"
timeout = "3s"
hostname = "myhost.com"
port = 8080
[backends.backend1.healthcheck.headers]

3
docs/configuration/backends/consulcatalog.md
View file

@ -106,7 +106,8 @@ Additional settings can be defined using Consul Catalog tags.
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `<prefix>.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend. ex: `NetworkErrorRatio() > 0.` |
| `<prefix>.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `<prefix>.backend.healthcheck.interval=1s` | Defines the health check interval. |
| `<prefix>.backend.healthcheck.interval=5s` | Defines the health check interval. |
| `<prefix>.backend.healthcheck.timeout=3s` | Defines the health check request timeout |
| `<prefix>.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `<prefix>.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

3
docs/configuration/backends/docker.md
View file

@ -223,7 +223,8 @@ Labels can be used on containers to override default behavior.
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `traefik.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend |
| `traefik.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.interval=1s` | Defines the health check interval. |
| `traefik.backend.healthcheck.interval=5s` | Defines the health check interval. |
| `traefik.backend.healthcheck.timeout=3s` | Defines the health check request timeout. |
| `traefik.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

3
docs/configuration/backends/ecs.md
View file

@ -145,7 +145,8 @@ Labels can be used on task containers to override default behavior:
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `traefik.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend |
| `traefik.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.interval=1s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.interval=5s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.timeout=3s` | Defines the health check request timeout. (Default: 5s) |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `traefik.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

1
docs/configuration/backends/file.md
View file

@ -37,6 +37,7 @@ Træfik can be configured with a file.
path = "/health"
port = 88
interval = "30s"
timeout = "5s"
scheme = "http"
hostname = "myhost.com"
[backends.backend1.healthcheck.headers]

3
docs/configuration/backends/marathon.md
View file

@ -209,7 +209,8 @@ The following labels can be defined on Marathon applications. They adjust the be
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `traefik.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend |
| `traefik.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.interval=1s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.interval=5s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.timeout=3s` | Defines the health check request timeout. (Default: 5s) |
| `traefik.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

3
docs/configuration/backends/mesos.md
View file

@ -123,7 +123,8 @@ The following labels can be defined on Mesos tasks. They adjust the behavior for
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `traefik.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend |
| `traefik.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.interval=1s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.interval=5s` | Defines the health check interval. (Default: 30s) |
| `traefik.backend.healthcheck.timeout=3s` | Defines the health check request timeout. (Default: 5s) |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `traefik.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

3
docs/configuration/backends/rancher.md
View file

@ -153,7 +153,8 @@ Labels can be used on task containers to override default behavior:
| `traefik.backend.buffering.retryExpression=EXPR` | See [buffering](/configuration/commons/#buffering) section. |
| `traefik.backend.circuitbreaker.expression=EXPR` | Creates a [circuit breaker](/basics/#backends) to be used against the backend |
| `traefik.backend.healthcheck.path=/health` | Enables health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.interval=1s` | Defines the health check interval. |
| `traefik.backend.healthcheck.interval=5s` | Defines the health check interval. |
| `traefik.backend.healthcheck.timeout=3s ` | Defines the health check request timeout. |
| `traefik.backend.healthcheck.port=8080` | Sets a different port for the health check. |
| `traefik.backend.healthcheck.scheme=http` | Overrides the server URL scheme. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Defines the health check hostname. |

3
docs/configuration/backends/servicefabric.md
View file

@ -103,7 +103,8 @@ Labels, set through extensions or the property manager, can be used on services
| `traefik.servicefabric.enablelabeloverrides` | Toggle whether labels can be overridden using the Service Fabric Property Manager API |
| `traefik.backend.healthcheck.path=/health` | Enable health check for the backend, hitting the container at `path`. |
| `traefik.backend.healthcheck.port=8080` | Allow to use a different port for the health check. |
| `traefik.backend.healthcheck.interval=1s` | Define the health check interval. |
| `traefik.backend.healthcheck.interval=5s` | Define the health check interval. |
| `traefik.backend.healthcheck.timeout=3s` | Define the health check request timeout. |
| `traefik.backend.healthcheck.hostname=foobar.com` | Define the health check hostname. |
| `traefik.backend.healthcheck.headers=EXPR` | Define the health check request headers <br>Format: <code>HEADER:value&vert;&vert;HEADER2:value2</code> |
| `traefik.backend.loadbalancer.method=drr` | Override the default `wrr` load balancer algorithm |

14
docs/configuration/commons.md
View file

@ -241,19 +241,23 @@ Example configuration:
# Enable custom health check options.
[healthcheck]
# Set the default health check interval.
# Set the default health check interval and timeout.
#
# Optional
# Default: "30s"
#
# interval = "30s"
# timeout = "5s"
```
- `interval` set the default health check interval.
Will only be effective if health check paths are defined.
Given provider-specific support, the value may be overridden on a per-backend basis.
Can be provided in a format supported by [time.ParseDuration](https://golang.org/pkg/time/#ParseDuration) or as raw values (digits).
- `interval` sets the default health check interval.
- `timeout` sets the default health check request timeout.
These options will only be effective if health check paths are defined.
Given provider-specific support, the value may be overridden on a per-backend basis.
Can be provided in a format supported by [time.ParseDuration](https://golang.org/pkg/time/#ParseDuration) or as raw values (digits).
If no units are provided, the value is parsed assuming seconds.
**Note:** the interval must be greater than the timeout. If configuration doesn't reflect this, the interval will be set to timeout + 1 second.
## Life Cycle

2
docs/user-guide/marathon.md
View file

@ -71,7 +71,7 @@ Beginning with version 1.4, Traefik respects readiness check results if the Trae
!!! note
Due to the way readiness check results are currently exposed by the Marathon API, ready tasks may be taken into rotation with a small delay.
It is on the order of one readiness check timeout interval (as configured on the application specifiation) and guarantees that non-ready tasks do not receive traffic prematurely.
It is on the order of one readiness check timeout interval (as configured on the application specification) and guarantees that non-ready tasks do not receive traffic prematurely.
If readiness checks are not possible, a current mitigation strategy is to enable [retries](/configuration/commons#retry-configuration) and make sure that a sufficient number of healthy application tasks exist so that one retry will likely hit one of those.
Apart from its probabilistic nature, the workaround comes at the price of increased latency.