Repository files navigation

Marathon-lb is a tool for managing HAProxy, by consumingMarathon's app state. HAProxy is afast, efficient, battle-tested, highly available load balancer with many advanced features which power a number of high-profile websites.

• Stateless design: no direct dependency on any third-party state store like ZooKeeper or etcd (except through Marathon)
• Idempotent and deterministic: scales horizontally
• Highly scalable:can achieve line-rate per instance, with multiple instances providing fault-tolerance and greater throughput
• Real-time LB updates, viaMarathon's event bus
• Support for Marathon'shealth checks
• Multi-cert TLS/SSL support
• Zero-downtime deployments
• Per-serviceHAProxy templates
• DC/OS integration
• Automated Docker image builds (mesosphere/marathon-lb)
• Global HAProxy templates which can be supplied at launch
• Supports IP-per-task integration, such asProject Calico
• Includestini zombies reaper

• Using marathon-lb
• Securing your service with TLS/SSL (blog post)

Take a look atthe marathon-lb wiki for example usage, templates, and more.

The marathon-lb scriptmarathon_lb.py connects to the marathon APIto retrieve all running apps, generates a HAProxy config and reloads HAProxy.By default, marathon-lb binds to the service port of every application andsends incoming requests to the application instances.

Services are exposed on their service port (seeService Discovery & Load Balancingfor reference) as defined in their Marathon definition. Furthermore, apps areonly exposed on LBs which have the same LB tag (or group) as defined in the Marathonapp's labels (usingHAPROXY_GROUP). HAProxy parameters can be tuned by specify labels in your app.

To create a virtual host or hosts theHAPROXY_{n}_VHOST label needs to be set on thegiven application. Applications with a vhost set will be exposed on ports 80and 443, in addition to their service port. Multiple virtual hosts may be specifiedinHAPROXY_{n}_VHOST using a comma as a delimiter between hostnames.

All applications are also exposed on port 9091, using theX-Marathon-App-IdHTTP header. See the documentation forHAPROXY_HTTP_FRONTEND_APPID_HEAD inthetemplates section

You can access the HAProxy statistics via:9090/haproxy?stats, and you canretrieve the current HAProxy config from the:9090/_haproxy_getconfig endpoint.

The package is currently availablefrom the universe.To deploy marathon-lb on the public slaves in your DC/OS cluster,simply run:

dcos package install marathon-lb

To configure a custom ssl-certificate, set the dcos cli optionssl-certto your concatenated cert and private key in .pem format. For more detailssee theHAProxy documentation.

For further customization, templates can be added by pointing the dcos clioptiontemplate-url to a tarball containing a directorytemplates/.Seecomments in script on how to name those.

Synopsis:docker run -e PORTS=$portnumber --net=host mesosphere/marathon-lb sse|poll ...

You must setPORTS environment variable to allow haproxy bind to this port.Syntax:docker run -e PORTS=9090 mesosphere/marathon-lb sse [other args]

You can pass in your own certificates for the SSL frontend by settingtheHAPROXY_SSL_CERT environment variable. If you need more than onecertificate you can specify additional ones by settingHAPROXY_SSL_CERT0 -HAPROXY_SSL_CERT100.

In SSE mode, the script connects to the marathon events endpoint to getnotified about state changes. This only works with Marathon 0.11.0 ornewer versions.

Syntax:docker run mesosphere/marathon-lb sse [other args]

If you can't use the HTTP callbacks, the script can poll the APIs to getthe schedulers state periodically.

Syntax:docker run mesosphere/marathon-lb poll [other args]

To change the poll interval (defaults to 60s), you can set thePOLL_INTERVALenvironment variable.

You can also run the update script directly.To generate an HAProxy configuration from Marathon running atlocalhost:8080 with themarathon_lb.py script, run:

$./marathon_lb.py --marathon http://localhost:8080 --group external --strict-mode --health-check

It is possible to pass--auth-credentials= option if your Marathon requires authentication:

$./marathon_lb.py --marathon http://localhost:8080 --auth-credentials=admin:password

It is possible to get the auth credentials (user & password) from VAULT if you define the followingenvironment variables before running marathon-lb: VAULT_TOKEN, VAULT_HOST, VAULT_PORT, VAULT_PATHwhere VAULT_PATH is the root path where your user and password are located.

This will refreshhaproxy.cfg, and if there were any changes, then it willautomatically reload HAProxy. Only apps with the labelHAPROXY_GROUP=externalwill be exposed on this LB.

marathon_lb.py has a lot of additional functionality like sticky sessions, HTTP to HTTPS redirection, SSL offloading, virtual host support and templating capabilities.

To get the full documentation run:

$./marathon_lb.py --help

You can provide your SSL certificate paths to be placed in frontend marathon_https_in section with--ssl-certs.

$./marathon_lb.py --marathon http://localhost:8080 --group external --ssl-certs /etc/ssl/site1.co,/etc/ssl/site2.co --health-check --strict-mode

If you are using the script directly, you have two options:

• Provide nothing and config will use/etc/ssl/cert.pem as the certificate path. Put the certificate in this path or edit the file for the correct path.
• Provide--ssl-certs command line argument and config will use these paths.

If you are using the providedrun script or Docker image, you have three options:

• Provide your certificate text inHAPROXY_SSL_CERT environment variable. Contents will be written to/etc/ssl/cert.pem. Config will use this path unless you specified extra certificate paths as in the next option.
• Provide SSL certificate paths with--ssl-certs command line argument. Your config will use these certificate paths.
• Provide nothing and it will create self-signed certificate on/etc/ssl/cert.pem and config will use it.

You can skip the configuration file validation (via calling HAProxy service) process if you don't have HAProxy installed. This is especially useful if you are running HAProxy on Docker containers.

$./marathon_lb.py --marathon http://localhost:8080 --group external --skip-validation

You can use HAProxy maps to speed up web application (vhosts) to backend lookup. This is very useful for large installations where the traditional vhost to backend rules comparison takes considerable time since it sequentially compares each rule. HAProxy map creates a hash based lookup table so its fast compared to the other approach, this is supported in marathon-lb using--haproxy-map flag.

$./marathon_lb.py --marathon http://localhost:8080 --group external --haproxy-map

Currently it creates a lookup dictionary only for host header (both HTTP and HTTPS) and X-Marathon-App-Id header. But for path based routing and auth, it uses the usual backend rules comparison.

Marathon-lb exposes a few endpoints on port 9090 (by default). They are:

* These endpoints won't function when marathon-lb is inpoll mode as there is no marathon-lb process to be signaled in this mode (marathon-lb exits after each poll).

App labels are specified in the Marathon app definition. These can be used to override HAProxy behaviour. For example, to specify theexternal group for an app with a virtual host namedservice.mesosphere.com:

{"id":"http-service","labels": {"HAPROXY_GROUP":"external","HAPROXY_0_VHOST":"service.mesosphere.com"  }}

Some labels are specifiedper service port. These are denoted with the{n} parameter in the label key, where{n} corresponds to the service port index, beginning at0.

Seethe configuration doc for the full listof labels.

Marathon-lb global templates (as listed in theLonghelp) can be overwritten in two ways:-By creating an environment variable in the marathon-lb container-By placing configuration files in thetemplates/ directory (relative to where the script is run from)

For example, to replaceHAPROXY_HTTPS_FRONTEND_HEAD with this content:

frontend new_frontend_label  bind *:443 ssl crt /etc/ssl/cert.pem  mode http

Then this environment variable could be added to the Marathon-LB configuration:

"HAPROXY_HTTPS_FRONTEND_HEAD": "\\nfrontend new_frontend_label\\n  bind *:443 ssl {sslCerts}\\n  mode http"

Alternately, a file calledHAPROXY_HTTPS_FRONTEND_HEAD could be placed intemplates/ directory through the use of an artifact URI.

Additionally, some templates can also beoverriddenper app service port. You may add your own templates to the Docker image, or provide them at startup.

Seethe configuration doc for the full list of templates.

Some templates may be overridden using app labels,as per thelabels section. Strings are interpreted as literalHAProxy configuration parameters, with substitutions respected (as per thetemplates section). The HAProxy configuration will be validatedfor correctness before reloading HAProxy after changes.Note: Since theHAProxy config is checked before reloading, if an app's HAProxylabels aren't syntactically correct, HAProxy will not be reloaded and mayresult in stale config.

Here is an example for a service calledhttp-service which requires thathttp-keep-alive be disabled:

{"id":"http-service","labels":{"HAPROXY_GROUP":"external","HAPROXY_0_BACKEND_HTTP_OPTIONS":"  option forwardfor\n  no option http-keep-alive\n  http-request set-header X-Forwarded-Port %[dst_port]\n  http-request add-header X-Forwarded-Proto https if { ssl_fc }\n"  }}

The full list of per service port templates which can be specifiedaredocumented here.

As a shortcut to add haproxy global default options (without overriding the global template) a comma-separatedlist of options may be specified via theHAPROXY_GLOBAL_DEFAULT_OPTIONS environment variable.The default value when not specified isredispatch,http-server-close,dontlognull; as an example, to add thehttplog option (and keep the existing defaults), one should specifyHAPROXY_GLOBAL_DEFAULT_OPTIONS=redispatch,http-server-close,dontlognull,httplog.

• Note that this setting has no effect when theHAPROXY_HEAD template has been overridden.

• Use service ports within the reserved range (which is 10000 to 10100 by default). This will prevent port conflicts, and ensure reloads don't result in connection errors.
• Avoid using theHAPROXY_{n}_PORT label; prefer defining service ports.
• Consider running multiple marathon-lb instances. In practice, 3 or more should be used to provide high availability for production workloads. Running 1 instance is never recommended, and unless you have significant load running more than 5 instances may not add value. The number of MLB instances you run will vary depending on workload and the amount of failure tolerance required. Note:do not run marathon-lb on every node in your cluster. This is considered an anti-pattern due to the implications of hammering the Marathon API and excess health checking.
• Consider using a dedicated load balancer in front of marathon-lb to permit upgrades/changes. Common choices include an ELB (on AWS) or a hardware load balancer for on-premise installations.
• Use separate marathon-lb groups (specified with--group) for internal and external load balancing. On DC/OS, the default group isexternal. A simpleoptions.json for an internal load balancer would be:

  {"marathon-lb": {"name":"marathon-lb-internal","haproxy-group":"internal","bind-http-https":false,"role":""    }  }

• For HTTP services, consider setting VHost (and optionally a path) to access the service on ports 80 and 443. Alternatively, the service can be accessed on port 9091 using theX-Marathon-App-Id header. For example, to access an app with the IDtweeter:

$ curl -vH "X-Marathon-App-Id: /tweeter" marathon-lb.marathon.mesos:9091/*   Trying 10.0.4.74...* Connected to marathon-lb.marathon.mesos (10.0.4.74) port 9091 (#0)> GET / HTTP/1.1> Host: marathon-lb.marathon.mesos:9091> User-Agent: curl/7.48.0> Accept: */*> X-Marathon-App-Id: /tweeter>< HTTP/1.1 200 OK

• Some of the features of marathon-lb assume that it is the only instance of itself running in a PID namespace. i.e. marathon-lb assumes that it is running in a container. Certain features like the/_mlb_signal endpoints and the/_haproxy_getpids endpoint (and by extension, zero-downtime deployments) may behave unexpectedly if more than one instance of marathon-lb is running in the same PID namespace or if there are other HAProxy processes in the same PID namespace.
• Sometimes it is desirable to get detailed container and HAProxy logging for easier debugging as well as viewing connection logging to frontends and backends. This can be achieved by setting theHAPROXY_SYSLOGD environment variable orcontainer-syslogd value inoptions.json like so:

  {"marathon-lb": {"container-syslogd":true    }  }

• Please note thatzdd.py is not to be used in a production environment and is purely developed for demonstration purposes.

Marathon-lb is able to perform canary style blue/green deployment with zero downtime. To execute such deployments, you must follow certain patterns when using Marathon.

The deployment method is describedin this Marathon document. Marathon-lb provides an implementation of the aforementioned deployment method with the scriptzdd.py. To perform a zero downtime deploy usingzdd.py, you must:

• Specify theHAPROXY_DEPLOYMENT_GROUP andHAPROXY_DEPLOYMENT_ALT_PORT labels in your app template
  • HAPROXY_DEPLOYMENT_GROUP: This label uniquely identifies a pair of apps belonging to a blue/green deployment, and will be used as the app name in the HAProxy configuration
  • HAPROXY_DEPLOYMENT_ALT_PORT: An alternate service port is required because Marathon requires service ports to be unique across all apps
• Only use 1 service port: multiple ports are not yet implemented
• Use the providedzdd.py script to orchestrate the deploy: the script will make API calls to Marathon, and use the HAProxy stats endpoint to gracefully terminate instances
• The marathon-lb container must be run in privileged mode (to executeiptables commands) due to the issues outlined in the excellent blog post by theYelp engineering team found here
• If you have long-lived TCP connections using the same HAProxy instances, it may cause the deploy to take longer than necessary. The script will wait up to 5 minutes (by default) for connections to drain from HAProxy between steps, but any long-lived TCP connections will cause old instances of HAProxy to stick around.

An example minimal configuration for atest instance of nginx is included here. You might execute a deployment from a CI tool like Jenkins with:

./zdd.py -j 1-nginx.json -m http://master.mesos:8080 -f -l http://marathon-lb.marathon.mesos:9090 --syslog-socket /dev/null

Zero downtime deployments are accomplished through the use of a Lua module, which reports the number of HAProxy processes which are currently running by hitting the stats endpoint at the/_haproxy_getpids. After a restart, there will be multiple HAProxy PIDs until all remaining connections have gracefully terminated. By waiting for all connections to complete, you may safely and deterministically drain tasks. A caveat of this, however, is that if you have any long-lived connections on the same LB, HAProxy will continue to run and serve those connections until they complete, thereby breaking this technique.

The ZDD script includes the ability to specify a pre-kill hook, which is executed before draining tasks are terminated. This allows you to run your own automated checks against the old and new app before the deploy continues.

Zdd has support to split the traffic between two versions of same app (version 'blue' and version 'green') by having instances of both versions live at the same time. This is supported with the help of theHAPROXY_DEPLOYMENT_NEW_INSTANCES label.

When you run zdd with the--new-instances flag, it creates only the specified number of instances of the new app, and deletes the same number of instances from the old app (instead of the normal, create all instances in new and delete all from old approach), to ensure that the number of instances in new app and old app together is equal toHAPROXY_DEPLOYMENT_TARGET_INSTANCES.

Example: Consider the same nginx app example where there are 10 instances of nginx running image version v1, now we can use zdd to create 2 instances of version v2, and retain 8 instances of V1 so that traffic is split in ratio 80:20 (old:new).

Creating 2 instances with new version automatically deletes 2 instances in existing version. You could do this using the following command:

$./zdd.py -j 1-nginx.json -m http://master.mesos:8080 -f -l http://marathon-lb.marathon.mesos:9090 --syslog-socket /dev/null --new-instances 2

This state where you have instances of both old and new versions of same app live at the same time is called hybrid state.

When a deployment group is in hybrid state, it needs to be converted into completely current version or completely previous version before deploying any further versions, this could be done with the help of the--complete-cur and--complete-prev flags in zdd.

When you run the below command, it converts all instances to new version so that traffic split ratio becomes 0:100 (old:new) and it deletes the old app. This is graceful as it follows usual zdd procedure of waiting for tasks/instances to drain before deleting them.

$./zdd.py -j 1-nginx.json -m http://master.mesos:8080 -f -l http://marathon-lb.marathon.mesos:9090 --syslog-socket /dev/null --complete-cur

Similarly you can use--complete-prev flag to convert all instances to old version (and this is essentially a rollback) so that traffic split ratio becomes 100:0 (old:new) and it deletes the new app.

Currently only one hop of traffic split is supported, so you can specify the number of new instances (directly proportional to traffic split ratio) only when app is having all instances of same version (completely blue or completely green). This implies--new-instances flag cannot be specified in hybrid mode to change traffic split ratio (instance ratio) as updating Marathon label (HAPROXY_DEPLOYMENT_NEW_INSTANCES) currently triggers new deployment in marathon which will not be graceful. Currently for the example mentioned, the traffic split ratio is 100:0 -> 80:20 -> 0:100, where there is only one hop when both versions get traffic simultaneously.

Marathon-lb supports load balancing for applications that use the Mesos IP-per-taskfeature, whereby each task is assigned unique, accessible, IP addresses. For thesetasks services are directly accessible via the configured discovery ports and thereis no host port mapping. Note, that due to limitations with Marathon (seemesosphere/marathon#3636)configured service ports are not exposed to marathon-lb for IP-per-task apps.

For these apps, if the service ports are missing from the Marathon app data,marathon-lb will automatically assign port values from a configurable range if youspecify it. The range is configured using the--min-serv-port-ip-per-task and--max-serv-port-ip-per-task options. While port assignment is deterministic, theassignment is not guaranteed if you change the current set of deployed apps. Inother words, when you deploy a new app, the port assignments may change.

When running with isolated containers, you may need to take care of reaping orphaned child processes. HAProxy typically produces orphan processes because of its two-step reload mechanism. Marathon-LB usestini for this purpose. When running in a container without PID namespace isolation, setting theTINI_SUBREAPER environment variable is recommended.

PRs are welcome, but here are a few general guidelines:

• Avoid making changes which may break existing behaviour

• Document new features

• Update/include tests for new functionality. To install dependencies and run tests:

  pip install -r requirements-dev.txtnosetests

• Use the pre-commit hook to automatically generate docs:

  bash /path/to/marathon-lb/scripts/install-git-hooks.sh

Running unit and integration tests is automated asmake targets. Dockeris required to use the targets as it will run all tests in containers.

Several environment variables can be set to control the image tags,DCOS version/variant, etc. Check the top of theMakefile for more info.

To run the unit tests:

make test-unit

To run the integration tests a DCOS installation will be started viadcos-e2e. The installation ofdcos-e2e and management of the cluster will all be done in dockercontainers. Since the installers are rather large downloads, it isbenificial to specify a value forDCOS_E2E_INSTALLERS_DIR. By defaultDCOS_E2E_INSTALLERS_DIR is inside the.cache directory that will beremoved uponmake clean. You must provide a repository for theresultant docker image to be pushed to via theCONTAINTER_REPOenvironemnt variable. It is assumed that the local docker is alreadylogged in and the image will be pushed prior to launching the cluster.

To run the integration tests on the OSS variant of DCOS:

DCOS_E2E_INSTALLERS_DIR="${HOME}/dcos/installers" \CONTAINTER_REPO="my_docker_user/my-marathon-lb-repo" make test-integration

To run the integration tests on the ENTERPRISE variant of DCOS:

DCOS_LICENSE_KEY_PATH=${HOME}/license.txt \DCOS_E2E_VARIANT=enterprise \DCOS_E2E_INSTALLERS_DIR="${HOME}/dcos/installers"\CONTAINTER_REPO="my_docker_user/my-marathon-lb-repo" make test-integration

To run both unit and integration tests (add appropriate variables):

CONTAINTER_REPO="my_docker_user/my-marathon-lb-repo" maketest

You need to install the curl development package.

# Fedoradnf install libcurl-devel# Ubuntuapt-get install libcurl-dev

Thepycurl package linked against the wrong SSL backend when you installed it.

pip uninstall pycurlexport PYCURL_SSL_LIBRARY=nsspip install -r requirements-dev.txt

Swapnss for whatever backend it mentions.

1. Create a Github release. Follow the convention of past releases. You can findsomething to copy/paste if you hit the "edit" button of a previous release.

2. The Github release creates a tag, and Dockerhub will build off of that tag.

3. Make a PR to Universe. The suggested way is to create one commit thatonly copiesthe previous dir to a new one, and then a second commit that makes the actual changes.If unsure, check out the previous commits to the marathon-lb directory in Universe.