Update multiple containers/stacks/services at once

[mircoianese]

Hello,

I'm using portainer to manage a server with about 80 stacks, each one with 3 containers. I'm now in a situation where I need to update the image of one of the containers, so I pulled the new image but I realized there not seems to be a quick way to redeploy all the stacks.

The only way I see is going to each stack and press the "Update the stack" button. It works, but doing that on 80 stacks is VERY slow.

Am I missing something?

I think it would be nice to have a Redeploy / Update button here:

Thanks

[papalouf]

Hello,

+1
For me, an Update button (to update the stack manifest and redeploy it) and a Redeploy button (to just pull latests images and redeploy the stack) should be a huge QoL enhancement when you manage many stacks.

I have to stop then start manually each stack that run not up to date images , it may be painful.

Thanks a lot to consider this enhancement.

[kikkoshoyu]

Would be extremely helpful. I tried just opening multiple tabs for each container and clicking through recreating and run into issues of the update process breaking partway through resulting in a bunch of "-old" containers.

[walgarch]

+1

Would love to see this feature!

Right now we have to click all the way into a container to get the "Recreate" button. Would be nice if we could select multiple containers on the "Container list" page and have that option there too to recreate (w/ the pull new image option)

[rsporsche]

+1

[creekmoon]

+1
Would love to see this feature!
It will be very helpful!

[martijndierckx]

+1 for both stacks (update) & containers (recreate)

[Monsterray]

Please add this. I just had to restore a backup of Portainer and have to go through each of my stacks and "Update the stack" on each of them.

[AdivonSlav]

Definitely a necessary feature, especially if running many containers

[carefulcomputer]

+1

[Cerulis]

+10

[cinorid]

+10

[deanfourie1]

Did this ever get implemented? I am looking to purchase portainer but need this function

[Asghan86]

+1
Feature still missing which is a pain in the ass!
If this does not change I will move to komo.do.

[ApachePig]

thanks for recommendation.

[Don-Swanson]

This can be accomplished via the API.

Here's how I accomplish this with python

import requests
import urllib3
import time
import os
from dotenv import load_dotenv

# Load environment variables
load_dotenv()

# Suppress only the single InsecureRequestWarning from urllib3
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

# Portainer credentials and URL from environment variables
portainer_url = os.getenv('PORTAINER_URL')
PAT = os.getenv('PORTAINER_PAT')

if not portainer_url or not PAT:
    raise ValueError("Missing required environment variables. Please check your .env file.")

# Authenticate and get JWT token
headers = {
    'X-API-Key': PAT
}

# Get the list of stacks
stacks_response = requests.get(f'{portainer_url}/stacks', headers=headers, verify=False)
stacks_response.raise_for_status()
stacks = stacks_response.json()


# Filter only active stacks (Status == 1)
active_stacks = [stack for stack in stacks if stack.get('Status') == 1]

# Update each active stack
for stack in active_stacks:
    stack_id = stack['Id']
    endpoint_id = stack['EndpointId']  # Get the endpoint ID for the stack

    print(f'Starting update of Stack {stack["Name"]}')
    start_time = time.time()  # Record the start time

    # Get the current stack details
    stack_details_response = requests.get(f'{portainer_url}/stacks/{stack_id}', headers=headers, verify=False)
    stack_details_response.raise_for_status()
    stack_details = stack_details_response.json()

    # Prepare the update payload
    update_payload = {
        'id': stack_id,
        'endpointId': endpoint_id,
        'pullImage': True,
        'env': stack_details.get('Env', [])
    }

    # Add StackFileContent if it exists
    if 'StackFileContent' in stack_details:
        update_payload['stackFileContent'] = stack_details['StackFileContent']
    else:
        # If StackFileContent is not directly available, try to retrieve it from another endpoint
        stack_file_response = requests.get(f'{portainer_url}/stacks/{stack_id}/file', headers=headers, verify=False)
        stack_file_response.raise_for_status()
        stack_file_content = stack_file_response.json().get('StackFileContent')
        if stack_file_content:
            update_payload['stackFileContent'] = stack_file_content

    # Update the stack
    update_response = requests.put(f'{portainer_url}/stacks/{stack_id}?endpointId={endpoint_id}', headers=headers, json=update_payload, verify=False)
    try:
        update_response.raise_for_status()
        end_time = time.time()  # Record the end time
        duration = end_time - start_time  # Calculate the duration
        print(f'Stack {stack["Name"]} updated successfully in {duration:.2f} seconds.')
    except requests.exceptions.HTTPError as err:
        print(f"HTTP error occurred: {err}")    

print('All active stacks updated.')

[SplinterHead]

I run Watchtower on each of my Portainer nodes.
On a schedule, it will check all stacks (unless excluded) for new images (by hash) and deploy the update

[Fnaargh]

I'm in the process of changing a driver, which requires me to redeploy everything, as such I'm forced to click into each stack and stop it, and the same again to redeploy.

Sure would've been swell to have a batch functionality at least for stopping/starting stacks just like we can with containers.
+1