How to safely shutdown ECS one node?

[Anonymous]

I can simply shutdown -h on the node, just don't know if any bad consequence could have. or we need to perform some docker stop before that?

service docker stop   ??  is this gracefully enough to  shutdown the service?

[msalem]

Hello Walker,

I will recommend to perform docker stop $(docker ps -a -q) to stop any running container then you can run shutdown -h. Please remember to re-map the mounting drivers after rebooting before starting any container.

Thanks,

Magdy

[Anonymous]

Hi Msalem,

Thanks for the information. So I need to,

1, docker stop $(docker ps -a -q)

2, service docker stop

3, shutdown -h

re-map the mounting drivers after reboot means need to re-mount the disks like below?

/dev/sdb1                250G  241G  9.8G  97% /ecs/uuid-1

/dev/dm-5                9.8G  1.5G  7.8G  16% /var/lib/docker/devicemapper/mnt/173ab4da4eec467ffe21c3f8fb4f6c31931321d5e7097d3a0154f21918790444

[msalem]

you do not need to stop the service. just docker stop and then shutdown.

For remap, you will need only re-map /dev/sdb1 to /ecs/uuid-1 or adding this command to you local startup bash to it will re-map automatic

Magdy

[JuanManuel1]

Hi.

Please, What is the command to re-map the disks?.  We have shutdown ECS, using this information, but now the container don't start properly...

[MikeAnderson1]

mount /dev/sdb1 /ecs/uuid-1

I have an AWS EC2 VM so it uses xvdc1:

sudo mount /dev/xvdc1 /ecs/uuid-1

You can use:

sudo fdisk -l

To find the device name. Use:

mount -l

To confirm that the mount was successful.