Sometimes the Chef API can't find nodes

[hltbra]

I don't know why it happens, but every week we have a few nodes that don't get deleted. The logs:

=Instance does not appear to be Chef Server managed.=

The code that prints that message is:

def handle(event, _context):
    """Lambda Handler"""
    log_event(event)

    # If you're using a self signed certificate change
    # the ssl_verify argument to False
    with chef.ChefAPI(CHEF_SERVER_URL, get_pem(), USERNAME, ssl_verify=VERIFY_SSL):
        instance_id = get_instance_id(event)
        try:
            search = chef.Search('node', 'ec2_instance_id:' + instance_id)
        except ChefServerNotFoundError as err:
            LOGGER.error(err)
            return False

        if len(search) != 0:
            for instance in search:
                node = chef.Node(instance.object.name)
                client = chef.Client(instance.object.name)
                try:
                    node.delete()
                    LOGGER.info('===Node Delete: SUCCESS===')
                    client.delete()
                    LOGGER.info('===Client Delete: SUCCESS===')
                    return True
                except ChefServerNotFoundError as err:
                    LOGGER.error(err)
                    return False
        else:
            LOGGER.info('=Instance does not appear to be Chef Server managed.=')
            return True

The above implies len(search) == 0, which means the node cannot be found. But it's there...

[hltbra]

I wonder if the event is sent early and only after that the node gets registered on Chef.