how to use secure docker registry(by CA) for mesos container? - mesos

In DCOS, I want to deploy a mesos container with a self-defined image which stored in a local secure docker registry, and it has been secured by CA (not username and password!)
The json is
{
"id": "/gpu-tflinker",
"cmd": "while [ true ] ; do nvidia-smi; sleep 5; done",
"cpus": 0.1,
"mem": 1024,
"gpus": 1,
"instances": 1,
"constraints": [
[
"hostname",
"CLUSTER",
"10.140.0.22"
]
],
"container": {
"type": "MESOS",
"docker": {
"image": "tflinker:test-gpu",
"credential": null
}
}
}
The above json failed to run on marathon, and there is no content on mesos's stderr and stdout file, on mesos-agent log, the error message is :
E0721 05:01:57.726367 22498 slave.cpp:3976] Container 'e2c68720-0fb7-41bc-9d3b-a2b5e4793816' for executor 'gpu-t
flinker.b6f96725-6dd1-11e7-ba5d-0242b2c758c0' of framework 1079aaea-6dde-4dc1-8990-d926a895de78-0000 failed to s
tart: Unexpected HTTP response '401 Unauthorized' when trying to get the manifest
W0721 05:01:57.726478 22497 composing.cpp:541] Container 'e2c68720-0fb7-41bc-9d3b-a2b5e4793816' is already destr
oyed
I0721 05:01:57.726583 22497 slave.cpp:4082] Executor 'gpu-tflinker.b6f96725-6dd1-11e7-ba5d-0242b2c758c0' of fram
ework 1079aaea-6dde-4dc1-8990-d926a895de78-0000 has terminated with unknown status
I0721 05:01:57.726603 22497 slave.cpp:4193] Cleaning up executor 'gpu-tflinker.b6f96725-6dd1-11e7-ba5d-0242b2c75
8c0' of framework 1079aaea-6dde-4dc1-8990-d926a895de78-0000
I0721 05:01:57.726794 22497 slave.cpp:4281] Cleaning up framework 1079aaea-6dde-4dc1-8990-d926a895de78-0000
so it seems mesos failed to fetch the docker image. I've configed CA file for dockerd(move ca files to /etc/docker/certs.d/), so I can 'docker pull' the image to local machine, but I am not sure how to config CA file for mesos~
in mesos-agent configurations, there exist a item --docker_config=VALUE, but it seems this item can only be used for username/password secured registry, I don't know how to config for CA secured registry.
anybody can help me out?! thanks!

I think CA file is just for encryption. you will need username and password in ca file way I think.
In my way, I put auth file into the container to authorize my private registry.
I wrote a web service for downloading the auth file
xxx.tar.gz(format: .docker/config.json in the tar.gz)
in the config.json, {"auths": {"test.com:6999": {"auth": "(username:password) [base64 encode]"}}} like {"auths": {"test.com:6999": {"auth": "Y2NjOjEyMw=="}}}
use Mesos uris to download auth files prepared into the containers. then, it would authorized.
"uris": [
"http:your download url"
]

Related

unable to control swarm ingress network with ansible

I'm deploying Docker swarm with ansible and I would like to ensure the ingress network has been created. In that aim, I configured the following task :
- name: Ensure ingress network exists
docker_network:
state: present
name: ingress
driver: overlay
driver_options:
ingress: true
And I'm getting the following error :
An exception occurred during task execution. To see the full traceback, use -vvv. The error was: docker.errors.NotFound: 404 Client Error for http+docker://localhost/v1.41/networks/ingress/disconnect: Not Found ("No such container: ingress-endpoint")
fatal: [swarm-srv-1]: FAILED! => {"changed": false, "msg": "An unexpected docker error occurred: 404 Client Error for http+docker://localhost/v1.41/networks/ingress/disconnect: Not Found (\"No such container: ingress-endpoint\")"}
I've tried to add some arguments likes :
scope: swarm
force: yes
But no changes... I've also tried to delete the ingress with ansible (state: absent), but I always get the same error.
Note that I don't face any issue when trying to delete a recreate the ingress network manually on the swarm : docker network rm ingress
I don't know how to resolve that issue...Any help would be appreciated. Thanks !
Here are some informations that may help...
# docker version
Version: 20.10.6
API version: 1.41
Go version: go1.13.15
Git commit: 370c289
Built: Fri Apr 9 22:47:35 2021
OS/Arch: linux/amd64
# docker inspect ingress
[
{
"Name": "ingress",
"Id": "yb2tkhep8vtaj9q7w3mssc9lx",
"Created": "2021-05-19T05:53:27.524446929-04:00",
"Scope": "swarm",
"Driver": "overlay",
"EnableIPv6": false,
"IPAM": {
"Driver": "default",
"Options": null,
"Config": [
{
"Subnet": "10.0.0.0/24",
"Gateway": "10.0.0.1"
}
]
},
"Internal": false,
"Attachable": false,
"Ingress": true,
"ConfigFrom": {
"Network": ""
},
"ConfigOnly": false,
"Containers": {
"ingress-sbox": {
"Name": "ingress-endpoint",
"EndpointID": "dfdc0f123d21a196c7a815c7e0a886924d0799ae5f3be2d38b64d527ed4620b1",
"MacAddress": "02:42:0a:00:00:02",
"IPv4Address": "10.0.0.2/24",
"IPv6Address": ""
}
},
"Options": {
"com.docker.network.driver.overlay.vxlanid_list": "4096"
},
"Labels": {},
"Peers": [
{
"Name": "8f8932d6f99f",
"IP": "(ip address here)"
},
{
"Name": "28b9ca95dcf0",
"IP": "(ip address here)"
},
{
"Name": "f7c48c8af2f5",
"IP": "(ip address here)"
}
]
}
]
I had the exact same issue when trying to customize the IP range of the ingress network. It looks like the docker_network module does not support modification of swarm specific networks: there is a open Github issue for this.
I went for the ugly workaround of removing the network by executing it through a shell (docker network rm ingress command) and adding it again. When adding it with the docker_network module, I found that adding also seems not be working (fails to set the ingress property of the network). So I ended up doing both remove- and create operation through a shell command.
Since the removal will trigger a confirmation dialogue:
WARNING! Before removing the routing-mesh network, make sure all the nodes in your swarm run the same docker engine version. Otherwise, removal may not be effective and functionality of newly create ingress networks will be impaired.
Are you sure you want to continue? [y/N]
I used the expect module to confirm the dialogue:
- name: remove default ingress network
ansible.builtin.expect:
command: docker network rm ingress
responses:
"[y/N]": "y"
- name: create customized ingress network
shell: "docker network create --ingress --subnet {{ docker_ingress_network }} --driver overlay ingress"
It is not perfect but it works.
There was one last problem I experienced: when running it on an existing swarm I ended up having network issues on the node where I did run this (somehow the docker_gwbridge network on that node could not handle the change). The fix for this was to fully remove the node and re-join the swarm (regenerates the docker_gwbridge).

PCF Config server Authentication issues

I am following the below to set up a config server in PCF and access the application in PCF.
https://github.com/pcf-guides/gs-configuration-server
I have added the config server from the PCF marketplace and it is bind with the app.
The config server is pointed to the following Git repository:
https://github.com/pcf-guides/configuration-server-config-repo.git
The VCAP_SERVICES is like below :
"p-config-server": [
{
"binding_name": null,
"credentials": {
"access_token_uri": "https://p-spring-cloud-
services.uaa.run.pivotal.io/oauth/token",
"client_id": "p-config-server-d837bb39-4cf1-47ce-994b-03257852a7f6",
"client_secret": "XXsWzX6IhKME",
"uri": "https://config-9f0e115f-dbb0-42e8-981a-e70e2cd62570.cfapps.io"
},
"instance_name": "sconfigserver",
"label": "p-config-server",
"name": "sconfigserver",
"plan": "trial",
"provider": null,
"syslog_drain_url": null,
"tags": [
"configuration",
"spring-cloud"
Whenever I am trying to access the application authentication error is coming :
Full authentication is required to access this resource.
I dont have any security configuration in the application.
In your application.properties file, you should add the following
management.security.enabled=false
This will disable security altogether.
I tried this in my main application class and worked fine for me
#SpringBootApplication(exclude = {SecurityAutoConfiguration.class , ManagementWebSecurityAutoConfiguration.class})

Querying remote registry service on machine <IP Address> resulted in exception: Unable to change open service manager

My cluster Config file as follows
`
{
"name": "SampleCluster",
"clusterConfigurationVersion": "1.0.0",
"apiVersion": "01-2017",
"nodes":
[
{
"nodeName": "vm0",
"iPAddress": "here is my VPS ip",
"nodeTypeRef": "NodeType0",
"faultDomain": "fd:/dc1/r0",
"upgradeDomain": "UD0"
},
{
"nodeName": "vm1",
"iPAddress": "here is my another VPS ip",
"nodeTypeRef": "NodeType0",
"faultDomain": "fd:/dc1/r1",
"upgradeDomain": "UD1"
},
{
"nodeName": "vm2",
"iPAddress": "here is my another VPS ip",
"nodeTypeRef": "NodeType0",
"faultDomain": "fd:/dc1/r2",
"upgradeDomain": "UD2"
}
],
"properties": {
"reliabilityLevel": "Bronze",
"diagnosticsStore":
{
"metadata": "Please replace the diagnostics file share with an actual file share accessible from all cluster machines.",
"dataDeletionAgeInDays": "7",
"storeType": "FileShare",
"IsEncrypted": "false",
"connectionstring": "c:\\ProgramData\\SF\\DiagnosticsStore"
},
"nodeTypes": [
{
"name": "NodeType0",
"clientConnectionEndpointPort": "19000",
"clusterConnectionEndpointPort": "19001",
"leaseDriverEndpointPort": "19002",
"serviceConnectionEndpointPort": "19003",
"httpGatewayEndpointPort": "19080",
"reverseProxyEndpointPort": "19081",
"applicationPorts": {
"startPort": "20001",
"endPort": "20031"
},
"isPrimary": true
}
],
"fabricSettings": [
{
"name": "Setup",
"parameters": [
{
"name": "FabricDataRoot",
"value": "C:\\ProgramData\\SF"
},
{
"name": "FabricLogRoot",
"value": "C:\\ProgramData\\SF\\Log"
}
]
}
]
}
}
It is almost identical to standalone service fabric download demo file for untrusted cluster except my VPS ip. I enabled remote registry service.I ran the
\TestConfiguration.ps1 -ClusterConfigFilePath \ClusterConfig.Unsecure.MultiMachine.json but i got the following error.
Unable to change open service manager handle because 5
Unable to query service configuration because System.InvalidOperationException: Unable to change open service manager ha
ndle because 5
at System.Fabric.FabricDeployer.FabricDeployerServiceController.GetServiceStartupType(String machineName, String serv
iceName)
Querying remote registry service on machine <IP Address> resulted in exception: Unable to change open service manager
handle because 5.
Unable to change open service manager handle because 5
Unable to query service configuration because System.InvalidOperationException: Unable to change open service manager ha
ndle because 5
at System.Fabric.FabricDeployer.FabricDeployerServiceController.GetServiceStartupType(String machineName, String serv
iceName)
Querying remote registry service on machine <Another IP Address> resulted in exception: Unable to change open service manager
handle because 5.
Best Practices Analyzer determined environment has an issue. Please see additional BPA log output in DeploymentTraces
LocalAdminPrivilege : True
IsJsonValid : True
IsCabValid :
RequiredPortsOpen : True
RemoteRegistryAvailable : False
FirewallAvailable :
RpcCheckPassed :
NoConflictingInstallations :
FabricInstallable :
DataDrivesAvailable :
Passed : False
Test Config failed with exception: System.InvalidOperationException: Best Practices Analyzer determined environment has
an issue. Please see additional BPA log output in DeploymentTraces folder.
at System.Management.Automation.MshCommandRuntime.ThrowTerminatingError(ErrorRecord errorRecord)
I don't understand the problem.VPSs are not locally connected. All are public IP.I don't know, this may b an issue. how do I make virtual LAN among these VPS?Can anyone give me some direction about this error?Anyone helps me is greatly appreciated.
Edit: I used VM term insted of VPS.
Finally I make this working. Actually all the nodes are in a network, i thought it wasn't. I enable file sharing. I try to access the shared file from the node where I ran configuration test to the all other nodes. I have to give the credentials of logins. And then it works like a charm.

Chronos can't run a private Docker container

I'm playing on localhost with a DC/OS installation. While everything works fine, I can't seem to run a docker image located inside a private repo. I'm using python to communicate with chronos:
#celery.task(name='add-job', soft_time_limit=5)
def add_job(job_id):
job_document = mongo.jobs.find_one({
'_id': job_id
})
if job_document:
worker_document = mongo.workers.find_one({
'_id': job_document['workerId']
})
if worker_document:
job = {
'async': True,
'name': job_document['_id'],
'owner': 'owner#gmail.com',
'command': "python /code/run.py",
"disabled": False,
"shell": True,
"cpus": worker_document['cpus'],
"disk": worker_document['disk'],
"mem": worker_document['memory'],
'schedule': 'R1//PT300S',# start now,
"epsilon": "PT60M",
"container": {
"type": "DOCKER",
"forcePullImage": True,
"image": "quay.io/username/container",
"network": "HOST",
"volumes": [{
"containerPath": "/images/",
"hostPath": "/images/",
"mode": "RW"
}]
},
"uris": [
"file:///images/docker.tar.gz"
]
}
return chronos_client.add(job)
else:
return 'worker not found'
else:
return 'job not found'
The job runs fine with a public image (alpine:latest) but it fails without any error inside the dcos installation.
The job gets executed but it fails immediately. The error log of the job inside chronos looks like this:
I1212 12:39:11.141639 25058 fetcher.cpp:498] Fetcher Info: {"cache_directory":"\/tmp\/mesos\/fetch\/slaves\/61d6d037-c9f5-482b-a441-11d85554461b-S1\/root","items":[{"action":"BYPASS_CACHE","uri":{"cache":false,"executable":false,"extract":false,"value":"file:\/\/\/images\/docker.tar.gz"}}],"sandbox_directory":"\/var\/lib\/mesos\/slave\/slaves\/61d6d037-c9f5-482b-a441-11d85554461b-S1\/docker\/links\/7029bbea-4c3d-439a-8720-411f6fe40eb9","user":"root"}
I1212 12:39:11.143575 25058 fetcher.cpp:409] Fetching URI 'file:///images/docker.tar.gz'
I1212 12:39:11.143587 25058 fetcher.cpp:250] Fetching directly into the sandbox directory
I1212 12:39:11.143602 25058 fetcher.cpp:187] Fetching URI 'file:///images/docker.tar.gz'
I1212 12:39:11.143612 25058 fetcher.cpp:167] Copying resource with command:cp '/images/docker.tar.gz' '/var/lib/mesos/slave/slaves/61d6d037-c9f5-482b-a441-11d85554461b-S1/docker/links/7029bbea-4c3d-439a-8720-411f6fe40eb9/docker.tar.gz'
I1212 12:39:11.146726 25058 fetcher.cpp:547] Fetched 'file:///images/docker.tar.gz' to '/var/lib/mesos/slave/slaves/61d6d037-c9f5-482b-a441-11d85554461b-S1/docker/links/7029bbea-4c3d-439a-8720-411f6fe40eb9/docker.tar.gz'
Stdout is empty. Executed directly inside marathon as an application with the same settings the authentication works and my image is downloaded & executed. Is this something that chronos does not support? It should...I mean, it has commands for docker...
Update: digging deeper into the agent logs I found this:
Failed to run 'docker -H unix:///var/run/docker.sock pull quay.io/username/container': exited with status 1; stderr='Error: Status 403 trying to pull repository username/container: "{\"error\": \"Permission Denied\"}"
I tried the archive with it's config.json file on the agent itself and it can download when triggered from the command line. I just can't seem to understand why chronos is not using it properly. I can't find any other reference on how to put my credentials other than this.
As it turns out...the uris param is deprecated in favor of fetch. I started from scratch with a marathon config applied to chronos and watched the logs carefully when I saw this: {'message': 'Tried to add both uri (deprecated) and fetch parameters on aBPepwhG5z33e4teG', 'status': 'Bad Request'}. Then I changed my uris parameter into:
"fetch": [{
"uri": "/images/docker.tar.gz",
"extract": true,
"executable": false,
"cache": false
}]
...and it worked.
your post looked a little like this one, which turned out to be a problem with volumes.

Play framework app heroku configuration error: Key 'application.conf' may not be followed by token: 'application.prod.conf'

I built an app based on template play-silhouette-seed-slick. template link
I got a configuration error caused by com.typesafe.config.ConfigException$Parse after deploying the app to heroku.
"Configuration error: Configuration error[ # file:/app/target/universal/stage/conf/: 2: Key 'application.conf' may not be followed by token: 'application.prod.conf' (if you intended 'application.prod.conf' to be part of a key or string value, try enclosing the key or value in double quotes)]"
The Procfile
web: target/universal/stage/bin/panobike-plus-server -Dhttp.port=${PORT} -Dconfig.resource=${PLAY_CONF_FILE}
And app.json
{
"name": "play-silhouette-slick-seed",
"description": "Seed project to show how Silhouette can be implemented into a Play Framework application with database access using Slick 3.",
"keywords": [
"Play",
"Silhouette",
"Slick"
],
"website": "https://github.com/sbrunk/play-silhouette-slick-seed",
"repository": "https://github.com/sbrunk/play-silhouette-slick-seed",
"success_url": "/",
"env": {
"BUILDPACK_URL": "https://github.com/heroku/heroku-buildpack-scala.git",
"PLAY_CONF_FILE": "application.prod.conf",
"PLAY_APP_SECRET": "changeme",
"FACEBOOK_CLIENT_ID": "",
"FACEBOOK_CLIENT_SECRET": "",
"GOOGLE_CLIENT_ID": "",
"GOOGLE_CLIENT_SECRET": "",
"TWITTER_CONSUMER_KEY": "",
"TWITTER_CONSUMER_SECRET": ""
}
}
In my production config "application.prod.conf", there is no such key "application.conf".
What does this error message mean?
Thank you
It was a stupid question.
I did not call the https://api.heroku.com/app-setups endpoint to setup the app.json enabled application on Heroku.
I had the same error. That was due to lack of PLAY_CONF_FILE env variable. To fix this error you need to open Heroku web page -> Settings -> click on Config Vars button and set new PLAY_CONF_FILE variable. For example application.staging.conf

Resources