Closed dqi-bloomberg closed 3 years ago
Can you verify the docker container used to launch the images? Somehow it seems to break the launch script for riot, leading to the first instances consuming all hugepages. I successfully launch 2 vmx instances using the same version (junos-vmx-x86-64-18.4R3.3.qcow2). Did you rebuild the image locally or simply renamed it?
mwiget@ryzen9:~/git/OpenJNPR-Container-vMX$ make ps
docker-compose ps
Name Command State Ports
--------------------------------------------------------------------------------------------------
openjnpr-container-vmx_vmx1_1 /launch.sh Up 0.0.0.0:32827->22/tcp, 0.0.0.0:32826->830/tcp
openjnpr-container-vmx_vmx2_1 /launch.sh Up 0.0.0.0:32831->22/tcp, 0.0.0.0:32830->830/tcp
docker-compose -f regression/docker-compose.yml ps
Name Command State Ports
------------------------------
./getpass.sh
vMX openjnpr-container-vmx_vmx1_1 (v4:172.27.0.2 v6:) 18.4R3.3 AeMuphuth4dee4paeyahGhuk ready
vMX openjnpr-container-vmx_vmx2_1 (v4:172.27.0.3 v6:) 18.4R3.3 ikie8Zuaqui2eShoob7aev3u ready
I'm using the following docker-compose.yml file:
$ cat docker-compose.yml
# Copyright (c) 2017, Juniper Networks, Inc.
# # All rights reserved.
#
version: "3"
services:
vmx1:
image: juniper/openjnpr-container-vmx:bionic
privileged: true
tty: true
stdin_open: true
ports:
- "22"
- "830"
environment:
- ID=vmx1
- LICENSE=license-eval.txt
- IMAGE=junos-vmx-x86-64-18.4R3.3.qcow2
- PUBLICKEY=id_rsa.pub
- CONFIG=vmx1.conf
- IO_CORE=2
- WORKER_CORE=3
- MASTER_CORE=4
- VMXT_CORE=5
volumes:
- $PWD:/u:ro
- /var/run/docker.sock:/var/run/docker.sock
networks:
mgmt:
net-a:
net-b:
net-c:
vmx2:
image: juniper/openjnpr-container-vmx:bionic
privileged: true
tty: true
stdin_open: true
ports:
- "22"
- "830"
environment:
- ID=vmx2
- LICENSE=license-eval.txt
- IMAGE=junos-vmx-x86-64-18.4R3.3.qcow2
- PUBLICKEY=id_rsa.pub
- CONFIG=vmx2.conf
- IO_CORE=6
- WORKER_CORE=7
- MASTER_CORE=8
- VMXT_CORE=9
volumes:
- $PWD:/u:ro
- /var/run/docker.sock:/var/run/docker.sock
networks:
mgmt:
net-a:
net-b:
net-c:
networks:
mgmt:
net-a:
net-b:
net-c:
Verify the version by pulling it (again) from docker hub:
$ docker pull juniper/openjnpr-container-vmx:bionic
bionic: Pulling from juniper/openjnpr-container-vmx
Digest: sha256:4e7b54d8342aaecc40e5085a3a79f530efcc76e760f80697c2ebd6043526914d
Status: Image is up to date for juniper/openjnpr-container-vmx:bionic
docker.io/juniper/openjnpr-container-vmx:bionic
And ran this on ubuntu 20.04.1:
mwiget@ryzen9:~/git/OpenJNPR-Container-vMX$ uname -a
Linux ryzen9 5.4.0-56-generic #62-Ubuntu SMP Mon Nov 23 19:20:19 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
mwiget@ryzen9:~/git/OpenJNPR-Container-vMX$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 20.04.1 LTS
Release: 20.04
Codename: focal
Hey, I am sorry. Really tied up all day. Thanks so much for the help. Yes. i rebuilt the images locally.
I somehow resolved the issue by comment out the specific cores for the 2nd vmx.
dqi@dev-10-42-3-111:~/work-dev/robot$ cat docker-compose.yml
# Copyright (c) 2017, Juniper Networks, Inc.
# # All rights reserved.
#
version: "3"
services:
robot1:
image: artprod.dev.bloomberg.com/nfun/docker-robot:latest
stdin_open: true
ports:
- "22"
environment:
- ID=robot1
- PUBLICKEY=id_rsa.pub
volumes:
- $PWD:/u
networks:
mgmt:
vmx1:
image: artprod.dev.bloomberg.com/nfun/docker-juniper-vmx:latest
privileged: true
tty: true
stdin_open: true
ports:
- "22"
- "830"
environment:
- ID=vmx1
- LICENSE=license-eval.txt
- IMAGE=junos-vmx-x86-64-18.4R3.3.qcow2
- PUBLICKEY=id_rsa.pub
- CONFIG=vmx1.conf
# - WAITFOR=eth4
- IO_CORE=2
- WORKER_CORE=3
- MASTER_CORE=4
# - VMXT_CORE=5
# - HDDIMAGE=/images/p1.qcow2 # if we want it to be persistant
volumes:
# - $PWD/images:/images
- $PWD:/u:ro
- /var/run/docker.sock:/var/run/docker.sock
networks:
mgmt:
net-a:
net-b:
net-c:
vmx2:
image: artprod.dev.bloomberg.com/nfun/docker-juniper-vmx:latest
privileged: true
tty: true
stdin_open: true
ports:
- "22"
- "830"
environment:
- ID=vmx2
- LICENSE=license-eval.txt
- IMAGE=junos-vmx-x86-64-18.4R3.3.qcow2
- PUBLICKEY=id_rsa.pub
- CONFIG=vmx2.conf
# - WAITFOR=eth4
# - IO_CORE=6
# - WORKER_CORE=7
# - MASTER_CORE=8
# - VMXT_CORE=9
# - HDDIMAGE=/images/p1.qcow2 # if we want it to be persistant
volumes:
# - $PWD/images:/images
- $PWD:/u:ro
- /var/run/docker.sock:/var/run/docker.sock
networks:
mgmt:
net-a:
net-b:
net-c:
networks:
mgmt:
net-a:
net-b:
net-c:
dqi@dev-10-42-3-111:~/work-dev/robot$ ./getpass.sh
vMX robot_vmx2_1 (v4:172.21.0.3 v6:) 18.4R3.3 tha9Ohv0Shachai2ieLeegho ready
vMX robot_vmx1_1 (v4:172.21.0.2 v6:) 18.4R3.3 bei0IChohcaePeek2Sha3quo ready
dqi@dev-10-42-3-111:~/work-dev/robot$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 18.04.5 LTS
Release: 18.04
Codename: bionic
dqi@dev-10-42-3-111:~/work-dev/robot$ uname -a
Linux dev-10-42-3-111 4.15.0-126-generic #129-Ubuntu SMP Mon Nov 23 18:53:38 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
Great it worked out.
Thanks again for the help! please close the issue. Have a great day:-)
I have two vmx configured in the docker-compose.yml but only one is up. I reconfigured huge memory to 4GB instead 2GB. Still the same result.