docker/swarm

Swarm: a Docker-native clustering system

Adam Jorgensen

@OOPMan

...oh the other, I'm not happy that it seems to rely on a seemingly trivial thing to work...

For reference, this is my compose file: https://pastebin.com/gAhMyDp8

The referenced config and shell script files referenced are:
vault.json: https://pastebin.com/wKzERqCp
vault.sh: https://pastebin.com/iYJUPe38

Any thoughts on what might be happening would be greatly appreciated

Adam Jorgensen

@OOPMan

Here's some extra stuff too:
Various outputs from working example: https://pastebin.com/2Y6P1jQh
The same outputs from failing example: https://pastebin.com/gL53xhau

When I run the network inspect output through a JSON diff tool the only differences are entirely expected

Mohamed Ait Taleb

@mohamedaittaleb

kmoutawakil@172.31.5.169

Angelos Bousias

@abousias

Hi, new to swarm and swarm mode, my question is: Can I do docker run against a swarm for one-off containers?

Mike Holloway

@mh720

Yes, you can run individual docker containers on nodes that are also in a swarm

Angelos Bousias

@abousias

Yes @mh720 , but this way I am not taking any advantage of the swarm. The point is to have multiple hosts but a single point to call docker run on.

My client, can only perform 'docker run' at the moment using the CLI

Ádám Liszkai

@adaliszk

Hmm, what kind of OS can restrict you to program arguments? :D

Angelos Bousias

@abousias

@adaliszk an older Jenkins server

Ádám Liszkai

@adaliszk

aaah, so you don't have access directly to the docker CLI

Angelos Bousias

@abousias

@adaliszk nope :(

Mike Holloway

@mh720

‘docker run’ knows nothing of other hosts. However, once multiple hosts are in a swarm, you could use ‘docker service’ with replicas set to 1 from a single manager node in the swarm to kick off containers that will be scheduled against any of the hosts in the swarm, or add constraints to direct them to a single host.

Ádám Liszkai

@adaliszk

@abousias What about connecting to the target server via SSH and then run the docker CLI command?

Mike Holloway

@mh720

You can also run a GUI for a swarm, such as swarmpit, which runs as a container on a manager node in a swarm and contacts swarmpit-agent containers running on each node to collect the present state of the swarm. Swarmpit is one of several GUIs available to manage docker swarms. Also take a look at https://github.com/swarmstack/swarmstack which includes installing swarmpit for you on your nodes.

Angelos Bousias

@abousias

@adaliszk not possible because it is done via a plugin, I'll have to break some stuff. So my only option to try the legacy mode

GUEST.it

@guestisp

Hi to all. Any idea on how to "share" a huge volume with multiple containers so that each container can't access the other container's shared space ? Something like a bind mount, where /mnt/hugevolume/container1 is mounted in container1 and /mnt/hugevolume/container2 is mounted in container2

Mike Holloway

@mh720

Just an update that https://github.com/swarmstack 0.8.0 has been released and includes ansible management of Docker swarm hosts, updates to the DevOps tool chain, and optional Docker Errbot images for alerting into social networks channels

Adam Jorgensen

@OOPMan

Hey guys, I've run into an interesting issue. I'm testing some Docker Swarm stuff using a trio of VirtualBox VMs on my system. Each VM is configured with 2 network adapters, the first being a host-only and the second being bridged. I have docker up and running on each VM and initialised a swarm on one node using docker swarm init --advertise-addr=192.168.56.102. However, when I try to join the swarm on my other nodes I get an error claiming

Error response from daemon: rpc error: code = Unavailable desc = all SubConns are in TransientFailure, latest connection error: connection error: desc = "transport: Error while dialing dial tcp 192.168.56.102:2377: connect: no route to host"

, which is funny because I can ping and ssh to the supposedly unavailable destination with no issues. Anyone got any ideas?

Adam Jorgensen

@OOPMan

Nvm, I'm a dummy, Firewall

Swastik Roy

@royswastik

Noob Question:
Hello everyone, I am trying to dockerize my production app(based on LAMP stack). Currenty, I am using plain git to manage code. I am trying to figure out

how docker plays with git. 2. Do I need to use Jenkins to always create new docker with the updated git code.

Ádám Liszkai

@adaliszk

The usual approach is to build a new image with the new code change. The image should only have what the application needs and not what a deployment need.

so the 2nd option is the way what's widely used 😉

Swastik Roy

@royswastik

What’s widely used? Using a Jenkins or something similar to hve a CI/CD in place?

Ádám Liszkai

@adaliszk

building an image with the updated code

it could be done with any CI/CD, not only Jenkins

Swastik Roy

@royswastik

True... thanks

Swastik Roy

@royswastik

If I dockerize my existing application with single node and two replicas, how do I maintain persistent user session in those two replicas?

Swastik Roy

@royswastik

I have looked at sticky session on docker docs (https://docs.docker.com/ee/ucp/interlock/usage/sessions/) . But, do I have to setup this everytime I dockerize an application which maintains user session?

Mike Holloway

@mh720

@royswastik check out https://github.com/swarmstack/caddy/blob/master/Dockerfile for an example Dockerfile using git to check out a project during a build. This image compresses the result in an intermediary image using UPX, then finally builds an image from scratch copying in the pre-build files. But the git example is likely what you’re after.

Swastik Roy

@royswastik

Thanks Mike..

Mike Holloway

@mh720

https://youtu.be/3FpTcVnvfRg

Mike Holloway

@mh720

Just added a docker-compose-singlebox.yml that can be used to test out the https://github.com/swarmstack/swarmstack DevOps tools without installing Portworx or other dedicated storage. Deploys on a mac running Docker in just minutes.

Swastik Roy

@royswastik

Hi I trying to implement sticky session on dockers-swarm with traefik, but I could not achieve session persistence over two replicas on same machine.
In my docker-compose.yml, I have added labels for traefik and added the loadbalancer as well. Am I missing anything?
Or is there a better way to maintain session persistence?

I have a PHP backend and taking about maintaining same PHP session variables across replicas

Swastik Roy

@royswastik

I have solved the above issue.
What are the options to monitor services/containers/nodes in docker swarm?

Anything simpler than promesthius, grafana?

Andrés Sánchez García

@asg1612

hi, @royswastik

I created a Ansible role. It deploy grafana and prometheus with traefik. It's a proof of concept.
https://github.com/asg1612/ansible-role-prom-grafana

Swastik Roy

@royswastik

@asg1612 Great thanks. I was able to bring prometheus up from command line, but not from docker-compose.yml . Thanks , your github is a good reference.

Andrés Sánchez García

@asg1612

@royswastik :-)

Mike Holloway

@mh720

swarmstack 0.9.8 has just been released, and replaces unsee with karma, adds Docker service healthchecks for many of the containers, added a SCALING section in the README.md, and added an InfluxDB compose file with documentation for both swarmstack and non-swarmstack users. YouTube vids

tcep01

@tcep01

Hi all, I have an issue with creating service templates (using 18.06.1 on CentOS). The following example from the Docu fails:

docker service create --name hosttempl --hostname="{{.Node.Hostname}}-{{.Node.ID}}-{{.Service.Name}}" busybox:latest top

If I remove the last part {{.Service.Name}} the service gets created, however the hostname shows literally as {{.Node.Hostname}}-{{.Node.ID}}

tcep01

@tcep01

The last part is working as it is, I was inspecting the service instead of the container, however why I can't append {{.Service.Name}} I still don't understand.

FaizalKhan

@smfaizalkhan

Hello All,
I have a docker swarm on runningon two different hosts ,Leader in Ubuntu and worker in Windows
on docker network ls i can see the network getting listed

C:\Users\Faizal>docker network ls
NETWORK ID NAME DRIVER SCOPE
18887aba757f bridge bridge local
00df062ded65 docker_default bridge local
1e42220f0b70 docker_gwbridge bridge local
90c53993c421 host host local
udozhr1kppne ingress overlay swarm
9e08a2fd21fe none null local
rtcs20mrhmnt overnet overlay swarm

Docker inspect gives the service myserice

docker network inspect overnet
[
{
"Name": "overnet",
"Id": "rtcs20mrhmntlmwh02upgk5f2",
"Created": "2019-01-02T06:41:23.5165731Z",
"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": false,
"ConfigFrom": {
"Network": ""
},
"ConfigOnly": false,
"Containers": {
"1e9a962147b2443d41537d98fd91a020d3aa090bf1ea2bacafb7ceacce9df99c": {
"Name": "myservice.2.ydxmr7tb8jlzu4g6jwvnc7ryr",
"EndpointID": "0bd35007868ff220064ddf3fa19f9e36708853b45bc6234635b3f11b45f9660e",
"MacAddress": "02:42:0a:00:00:24",
"IPv4Address": "10.0.0.36/24",
"IPv6Address": ""
},
"lb-overnet": {
"Name": "overnet-endpoint",
"EndpointID": "7ed5fc49f3846b6b80a30a48b43170f7225b2a2210e99ee2219ac14fed6a6182",
"MacAddress": "02:42:0a:00:00:25",
"IPv4Address": "10.0.0.37/24",
"IPv6Address": ""
}
},
"Options": {
"com.docker.network.driver.overlay.vxlanid_list": "4097"
},
"Labels": {},
"Peers": [
{
"Name": "eb0e60ee3298",
"IP": "192.168.0.8"
},
{
"Name": "8b49e41d1df7",
"IP": "192.168.0.7"
}
]
}
]

now i enter into conatiner and try to ping another conatiner on ubuntu ,but i dont get inof ,all the data is lost.

ping 10.0.0.36
PING 10.0.0.36 (10.0.0.36): 56 data bytes
64 bytes from 10.0.0.36: seq=0 ttl=64 time=0.068 ms
64 bytes from 10.0.0.36: seq=1 ttl=64 time=0.176 ms
64 bytes from 10.0.0.36: seq=2 ttl=64 time=0.185 ms
64 bytes from 10.0.0.36: seq=3 ttl=64 time=0.176 ms
^C
--- 10.0.0.36 ping statistics ---
4 packets transmitted, 4 packets received, 0% packet loss
round-trip min/avg/max = 0.068/0.151/0.185 ms
/ # ping 10.0.0.35
PING 10.0.0.35 (10.0.0.35): 56 data bytes
^C
--- 10.0.0.35 ping statistics ---
32 packets transmitted, 0 packets received, 100% packet loss

Help needed?

vishnuavenudev

@vishnuavenudev

@vishnuavenudev
RUN ln -sf /dev/stdout /var/log/myapp/services-info.log \
&& ln -sf /dev/stderr /var/log/myapp/services-error.log
i added this in docker
i am running it on aws ecs
i am getting the logs on cloudwatch but cannot tail inside container

tailf /var/log/myapp/services-info.log
tailf /dev/stdout
both of them not running
not showing any logs
base image for this docker is ubuntu

Where communities thrive

People

Repo info

Activity