[IMP] runbot: pull docker image only when needed

When a lot of Docker images are updated at the same time, all runbot
hosts will try to pull them at the same moment.

With this commit, only the images marked as `always_pull` will be pulled
and if one of them takes too much time to be pulled, we let the host
make another loop turn before continuing the images pull.

Finally, if a host uses the Docker registry, it will pull the remote
image when running a step, that way the image will be pulled
automatically if needed.

runbot/container.py

@@ -166,7 +166,7 @@ def _docker_pull(image_tag):
     """Pull a docker image from a registry.
     :param image_tag: the full image tag, including the registry host
     e.g.: `dockerhub.runbot102.odoo.com/odoo:PureNobleTest`
-    :return: tuple(success, image) where success is a boolean and image a Docker image object or None in case of failure
+    :return: DockerMager.result dict
     """
     with DockerManager(image_tag) as dm:
         for chunk in dm.consume(dm.docker_client.api.pull(image_tag, stream=True)):

runbot/docker_manager.py

@@ -57,6 +57,7 @@ class DockerManager:
     def __exit__(self, exception_type, exception_value, exception_traceback):
         if self.log_progress:
             _logger.info('Finished in %.2fs', self.duration)
+            self.result['log_progress'] = self.log_progress
         if exception_value:
             self.result['success'] = False
             _logger.warning(exception_value)

runbot/models/build.py

@@ -16,7 +16,7 @@ from psycopg2 import sql
 from psycopg2.extensions import TransactionRollbackError
 
 from ..common import dt2time, now, grep, local_pgadmin_cursor, s2human, dest_reg, os, list_local_dbs, pseudo_markdown, RunbotException, findall, sanitize, markdown_escape
-from ..container import docker_stop, docker_state, Command, docker_run
+from ..container import docker_stop, docker_state, Command, docker_run, docker_pull
 from ..fields import JsonDictField
 
 from odoo import models, fields, api
@@ -853,6 +853,15 @@ class BuildResult(models.Model):
         if 'image_tag' not in kwargs:
             kwargs.update({'image_tag': self.params_id.dockerfile_id.image_tag})
         self._log('Preparing', 'Using Dockerfile Tag [%s](/runbot/dockerfile/tag/%s)', kwargs['image_tag'], kwargs['image_tag'], log_type='markdown')
+        icp = self.env['ir.config_parameter']
+        docker_registry_host = self.env['runbot.host'].browse(int(icp.get_param('runbot.docker_registry_host_id', default=0)))
+        if docker_registry_host and self.host_id.use_remote_docker_registry:
+            result = docker_pull(f"dockerhub.{docker_registry_host.name}/{kwargs['image_tag']}")
+            if result['success']:
+                result['image'].tag(kwargs['image_tag'])
+            if result.get('log_progress'):
+                self._log('Docker Run', f'Docker image was pulled {"" if result["success"] else "with errors"}')
+
         containers_memory_limit = self.env['ir.config_parameter'].sudo().get_param('runbot.runbot_containers_memory', 0)
         if containers_memory_limit and 'memory' not in kwargs:
             kwargs['memory'] = int(float(containers_memory_limit) * 1024 ** 3)

runbot/models/docker.py

@@ -128,6 +128,7 @@ class Dockerfile(models.Model):
     arch_base = fields.Text(related='template_id.arch_base', readonly=False, related_sudo=True)
     dockerfile = fields.Text(compute='_compute_dockerfile', tracking=True)
     to_build = fields.Boolean('To Build', help='Build Dockerfile. Check this when the Dockerfile is ready.', default=False)
+    always_pull = fields.Boolean('Always pull', help='Always Pull on the hosts, not only at the use time', default=False, tracking=True)
     version_ids = fields.One2many('runbot.version', 'dockerfile_id', string='Versions')
     description = fields.Text('Description')
     view_ids = fields.Many2many('ir.ui.view', compute='_compute_view_ids', groups="runbot.group_runbot_admin")

runbot/models/host.py

@@ -130,11 +130,16 @@ class Host(models.Model):
         all_tags = set(all_docker_files.mapped('image_tag'))
         if docker_registry_host and self.use_remote_docker_registry and not is_registry:
             _logger.info('Pulling docker images...')
-            for dockerfile in all_docker_files:
+            total_duration = 0
+            for dockerfile in all_docker_files.filtered('always_pull'):
                 remote_tag = f'dockerhub.{docker_registry_host.name}/{dockerfile.image_tag}'
                 result = docker_pull(remote_tag)
                 if result['success']:
                     result['image'].tag(dockerfile.image_tag)
+                total_duration += result['duration']
+                if total_duration > 60:
+                    _logger.warning("Pulling images took more than 60 seconds... will continue later")
+                    break
         else:
             _logger.info('Building docker images...')
             for dockerfile in self.env['runbot.dockerfile'].search([('to_build', '=', True)]):

runbot/views/dockerfile_views.xml

@@ -13,6 +13,7 @@
                 <field name="name"/>
                 <field name="image_tag"/>
                 <field name="to_build"/>
+                <field name="always_pull"/>
                 <field name="version_ids" widget="many2many_tags"/>
                 <field name="project_ids" widget="many2many_tags"/>
                 <field name="template_id"/>
@@ -104,6 +105,8 @@
             <field name="image_tag"/>
             <field name="to_build" groups="!runbot.group_runbot_admin"/>
             <field name="to_build" widget="boolean_toggle" groups="runbot.group_runbot_admin"/>
+            <field name="always_pull" groups="!runbot.group_runbot_admin"/>
+            <field name="always_pull" widget="boolean_toggle" groups="runbot.group_runbot_admin"/>
             <field name="version_ids" widget="many2many_tags"/>
             <field name="project_ids" widget="many2many_tags"/>
             <field name="use_count"/>