ipxe: add --timeout parameter to kernel and initrd

In case of network issue during the download of the ramdisk or the
kernel, ipxe can freeze. The default timeout value is "unlimited" [0] [1].
If force_power_state_during_sync is False, ironic will just ignore the
node until someone do an human intervention.

By setting a timeout, we ensure ipxe will just give up. Depending on the
BIOS configuration, a new boot attempt may be done immedialey without the
hassle of a hard reboot.

https://bugzilla.redhat.com/show_bug.cgi?id=1310778

[0] http://ipxe.org/cmd/kernel
[1] http://lists.ipxe.org/pipermail/ipxe-devel/2014-October/003829.html

Closes-Bug: 1550417
Change-Id: I472dfb73044df50849c9cf72de90e59151698376

etc/ironic/ironic.conf.sample

@@ -1943,6 +1943,10 @@
 # file. (string value)
 #ipxe_boot_script=$pybasedir/drivers/modules/boot.ipxe
 
+# Timeout value (in seconds) for downloading an image via
+# iPXE. Defaults to 0 (no timeout) (integer value)
+#ipxe_timeout=0
+
 # The IP version that will be used for PXE booting. Can be
 # either 4 or 6. Defaults to 4. EXPERIMENTAL (string value)
 #ip_version=4

ironic/drivers/modules/ipxe_config.template

@@ -5,14 +5,14 @@ dhcp
 goto deploy
 
 :deploy
-kernel {{ pxe_options.deployment_aki_path }} selinux=0 disk={{ pxe_options.disk }} iscsi_target_iqn={{ pxe_options.iscsi_target_iqn }} deployment_id={{ pxe_options.deployment_id }} deployment_key={{ pxe_options.deployment_key }} ironic_api_url={{ pxe_options.ironic_api_url }} troubleshoot=0 text {{ pxe_options.pxe_append_params|default("", true) }} boot_option={{ pxe_options.boot_option }} ip=${ip}:${next-server}:${gateway}:${netmask} BOOTIF=${mac} {% if pxe_options.root_device %}root_device={{ pxe_options.root_device }}{% endif %} ipa-api-url={{ pxe_options['ipa-api-url'] }} ipa-driver-name={{ pxe_options['ipa-driver-name'] }} boot_mode={{ pxe_options['boot_mode'] }} initrd=deploy_ramdisk coreos.configdrive=0
+kernel {% if pxe_options.ipxe_timeout > 0 %}--timeout {{ pxe_options.ipxe_timeout }} {% endif %}{{ pxe_options.deployment_aki_path }} selinux=0 disk={{ pxe_options.disk }} iscsi_target_iqn={{ pxe_options.iscsi_target_iqn }} deployment_id={{ pxe_options.deployment_id }} deployment_key={{ pxe_options.deployment_key }} ironic_api_url={{ pxe_options.ironic_api_url }} troubleshoot=0 text {{ pxe_options.pxe_append_params|default("", true) }} boot_option={{ pxe_options.boot_option }} ip=${ip}:${next-server}:${gateway}:${netmask} BOOTIF=${mac} {% if pxe_options.root_device %}root_device={{ pxe_options.root_device }}{% endif %} ipa-api-url={{ pxe_options['ipa-api-url'] }} ipa-driver-name={{ pxe_options['ipa-driver-name'] }} boot_mode={{ pxe_options['boot_mode'] }} initrd=deploy_ramdisk coreos.configdrive=0
 
-initrd {{ pxe_options.deployment_ari_path }}
+initrd {% if pxe_options.ipxe_timeout > 0 %}--timeout {{ pxe_options.ipxe_timeout }} {% endif %}{{ pxe_options.deployment_ari_path }}
 boot
 
 :boot_partition
-kernel {{ pxe_options.aki_path }} root={{ ROOT }} ro text {{ pxe_options.pxe_append_params|default("", true) }} initrd=ramdisk
+kernel {% if pxe_options.ipxe_timeout > 0 %}--timeout {{ pxe_options.ipxe_timeout }} {% endif %}{{ pxe_options.aki_path }} root={{ ROOT }} ro text {{ pxe_options.pxe_append_params|default("", true) }} initrd=ramdisk
-initrd {{ pxe_options.ari_path }}
+initrd {% if pxe_options.ipxe_timeout > 0 %}--timeout {{ pxe_options.ipxe_timeout }} {% endif %}{{ pxe_options.ari_path }}
 boot
 
 :boot_whole_disk

ironic/drivers/modules/pxe.py

@@ -81,6 +81,10 @@ pxe_opts = [
                    'drivers/modules/boot.ipxe'),
                help=_('On ironic-conductor node, the path to the main iPXE '
                       'script file.')),
+    cfg.IntOpt('ipxe_timeout',
+               default=0,
+               help=_('Timeout value (in seconds) for downloading an image '
+                      'via iPXE. Defaults to 0 (no timeout)')),
     cfg.StrOpt('ip_version',
                default='4',
                choices=['4', '6'],
@@ -274,7 +278,8 @@ def _build_pxe_config_options(task, pxe_info):
         'pxe_append_params': _get_pxe_conf_option(task, 'pxe_append_params'),
         'tftp_server': CONF.pxe.tftp_server,
         'aki_path': kernel,
-        'ari_path': ramdisk
+        'ari_path': ramdisk,
+        'ipxe_timeout': CONF.pxe.ipxe_timeout * 1000
     }
 
     return pxe_options

ironic/tests/unit/common/test_pxe_utils.py

@@ -45,6 +45,7 @@ class TestPXEUtils(db_base.DbTestCase):
                                    u'f33c123/deploy_ramdisk',
             'root_device': 'vendor=fake,size=123',
             'ipa-api-url': 'http://192.168.122.184:6385',
+            'ipxe_timeout': 0,
         }
 
         self.pxe_options = {
@@ -89,6 +90,11 @@ class TestPXEUtils(db_base.DbTestCase):
         }
         self.ipxe_options_bios.update(self.ipxe_options)
 
+        self.ipxe_options_timeout = self.ipxe_options_bios.copy()
+        self.ipxe_options_timeout.update({
+            'ipxe_timeout': 120
+        })
+
         self.ipxe_options_uefi = {
             'boot_mode': 'uefi',
         }
@@ -137,6 +143,25 @@ class TestPXEUtils(db_base.DbTestCase):
 
         self.assertEqual(six.text_type(expected_template), rendered_template)
 
+    def test__build_ipxe_timeout_config(self):
+        # NOTE(lucasagomes): iPXE is just an extension of the PXE driver,
+        # it doesn't have it's own configuration option for template.
+        # More info:
+        # http://docs.openstack.org/developer/ironic/deploy/install-guide.html
+        self.config(
+            pxe_config_template='ironic/drivers/modules/ipxe_config.template',
+            group='pxe'
+        )
+        self.config(http_url='http://1.2.3.4:1234', group='deploy')
+        rendered_template = pxe_utils._build_pxe_config(
+            self.ipxe_options_timeout, CONF.pxe.pxe_config_template,
+            '{{ ROOT }}', '{{ DISK_IDENTIFIER }}')
+
+        tpl_file = 'ironic/tests/unit/drivers/ipxe_config_timeout.template'
+        expected_template = open(tpl_file).read().rstrip()
+
+        self.assertEqual(six.text_type(expected_template), rendered_template)
+
     def test__build_ipxe_uefi_config(self):
         # NOTE(lucasagomes): iPXE is just an extension of the PXE driver,
         # it doesn't have it's own configuration option for template.

ironic/tests/unit/drivers/ipxe_config_timeout.template

@@ -0,0 +1,19 @@
+#!ipxe
+
+dhcp
+
+goto deploy
+
+:deploy
+kernel --timeout 120 http://1.2.3.4:1234/deploy_kernel selinux=0 disk=cciss/c0d0,sda,hda,vda iscsi_target_iqn=iqn-1be26c0b-03f2-4d2e-ae87-c02d7f33c123 deployment_id=1be26c0b-03f2-4d2e-ae87-c02d7f33c123 deployment_key=0123456789ABCDEFGHIJKLMNOPQRSTUV ironic_api_url=http://192.168.122.184:6385 troubleshoot=0 text test_param boot_option=netboot ip=${ip}:${next-server}:${gateway}:${netmask} BOOTIF=${mac} root_device=vendor=fake,size=123 ipa-api-url=http://192.168.122.184:6385 ipa-driver-name=pxe_ssh boot_mode=bios initrd=deploy_ramdisk coreos.configdrive=0
+
+initrd --timeout 120 http://1.2.3.4:1234/deploy_ramdisk
+boot
+
+:boot_partition
+kernel --timeout 120 http://1.2.3.4:1234/kernel root={{ ROOT }} ro text test_param initrd=ramdisk
+initrd --timeout 120 http://1.2.3.4:1234/ramdisk
+boot
+
+:boot_whole_disk
+sanboot --no-describe

ironic/tests/unit/drivers/modules/test_pxe.py

@@ -183,11 +183,13 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
     @mock.patch.object(pxe_utils, '_build_pxe_config', autospec=True)
     def _test_build_pxe_config_options(self, build_pxe_mock,
                                        whle_dsk_img=False,
-                                       ipxe_enabled=False):
+                                       ipxe_enabled=False,
+                                       ipxe_timeout=0):
         self.config(pxe_append_params='test_param', group='pxe')
         # NOTE: right '/' should be removed from url string
         self.config(api_url='http://192.168.122.184:6385', group='conductor')
         self.config(disk_devices='sda', group='pxe')
+        self.config(ipxe_timeout=ipxe_timeout, group='pxe')
 
         driver_internal_info = self.node.driver_internal_info
         driver_internal_info['is_whole_disk_image'] = whle_dsk_img
@@ -223,6 +225,8 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
             ramdisk = 'no_ramdisk'
             kernel = 'no_kernel'
 
+        ipxe_timeout_in_ms = ipxe_timeout * 1000
+
         expected_options = {
             'ari_path': ramdisk,
             'deployment_ari_path': deploy_ramdisk,
@@ -230,6 +234,7 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
             'aki_path': kernel,
             'deployment_aki_path': deploy_kernel,
             'tftp_server': tftp_server,
+            'ipxe_timeout': ipxe_timeout_in_ms,
         }
 
         image_info = {'deploy_kernel': ('deploy_kernel',
@@ -268,6 +273,11 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
         self._test_build_pxe_config_options(whle_dsk_img=False,
                                             ipxe_enabled=False)
 
+    def test__build_pxe_config_options_ipxe_and_ipxe_timeout(self):
+        self._test_build_pxe_config_options(whle_dsk_img=True,
+                                            ipxe_enabled=True,
+                                            ipxe_timeout=120)
+
     @mock.patch.object(pxe_utils, '_build_pxe_config', autospec=True)
     def test__build_pxe_config_options_whole_disk_image(self,
                                                         build_pxe_mock,
@@ -303,6 +313,7 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
             'tftp_server': tftp_server,
             'aki_path': 'no_kernel',
             'ari_path': 'no_ramdisk',
+            'ipxe_timeout': 0,
         }
 
         image_info = {'deploy_kernel': ('deploy_kernel',
@@ -345,7 +356,8 @@ class PXEPrivateMethodsTestCase(db_base.DbTestCase):
             'pxe_append_params': 'my-pxe-append-params',
             'tftp_server': 'my-tftp-server',
             'aki_path': 'no_kernel',
-            'ari_path': 'no_ramdisk'}
+            'ari_path': 'no_ramdisk',
+            'ipxe_timeout': 0}
         self.assertEqual(expected_options, options)
 
     @mock.patch.object(deploy_utils, 'fetch_images', autospec=True)

releasenotes/notes/ipxe_timeout_parameter-03fc3c76c520fac2.yaml

@@ -0,0 +1,4 @@
+---
+features:
+  - Add the ability to adjust ipxe timeout during image downloading, default is
+    still unlimited (0).