openstack/ironic
Code Issues Proposed changes

Retry stdlib ssl.SSLError

Browse Source 
Apparently, connection to the agent can fail with

    Unexpected error when processing next clean step.
    SSLError: ('timed out',): ssl.SSLError: ('timed out',)

This change adds SSLError to retriable errors since most of them are
related to temporary conditions, and the certificate validation failure
is handled by requests separately.

Change-Id: I4805d410b4bc98196e77d8c44a217e363dc88f0a
This commit is contained in:
Dmitry Tantsur 2021-08-05 16:17:03 +02:00
parent 512364df9d
commit 528eb9e4a9
3 changed files with 13 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
ironic/drivers/modules/agent_client.py
View File

@ -14,6 +14,7 @@
from http import client as http_client from http import client as http_client
import os import os
import ssl
from ironic_lib import metrics_utils from ironic_lib import metrics_utils
from oslo_log import log from oslo_log import log
@ -203,7 +204,7 @@ class AgentClient(object):
url, params=request_params, data=body, url, params=request_params, data=body,
verify=self._get_verify(node), verify=self._get_verify(node),
timeout=CONF.agent.command_timeout) timeout=CONF.agent.command_timeout)
except (requests.ConnectionError, requests.Timeout) as e: except (requests.ConnectionError, requests.Timeout, ssl.SSLError) as e:
result = self._handle_timeout_on_command_execution(node, method, result = self._handle_timeout_on_command_execution(node, method,
params, e) params, e)
response = None response = None

12
ironic/tests/unit/drivers/modules/test_agent_client.py
View File

@ -14,6 +14,7 @@
from http import client as http_client from http import client as http_client
import json import json
import ssl
from unittest import mock from unittest import mock
import requests import requests
@ -757,10 +758,11 @@ class TestAgentClientAttempts(base.TestCase):
method = 'standby.run_image' method = 'standby.run_image'
image_info = {'image_id': 'test_image'} image_info = {'image_id': 'test_image'}
params = {'image_info': image_info} params = {'image_info': image_info}
self.client.session.post.side_effect = [requests.Timeout(error), self.client.session.post.side_effect = [
requests.Timeout(error), requests.Timeout(error),
requests.Timeout(error), ssl.SSLError('timed out'),
requests.Timeout(error)] requests.ConnectionError(error),
requests.Timeout(error)]
self.client._get_command_url(self.node) self.client._get_command_url(self.node)
self.client._get_command_body(method, params) self.client._get_command_body(method, params)
@ -781,7 +783,7 @@ class TestAgentClientAttempts(base.TestCase):
image_info = {'image_id': 'test_image'} image_info = {'image_id': 'test_image'}
params = {'image_info': image_info} params = {'image_info': image_info}
self.client.session.post.side_effect = [requests.Timeout(error), self.client.session.post.side_effect = [requests.Timeout(error),
requests.Timeout(error), ssl.SSLError('timed out'),
MockResponse(response_data)] MockResponse(response_data)]
response = self.client._command(self.node, method, params) response = self.client._command(self.node, method, params)

4
releasenotes/notes/sslerror-287edf7f8b3c5f1c.yaml Normal file
View File

@ -0,0 +1,4 @@
---
fixes:
- |
Retries ``ssl.SSLError`` when connecting to the agent.