linux/tools/testing/selftests/tc-testing/plugin-lib/nsPlugin.py

252 lines
7.7 KiB
Python
Raw Permalink Normal View History

import os
import signal
from string import Template
import subprocess
import time
from multiprocessing import Pool
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
from functools import cached_property
from TdcPlugin import TdcPlugin
from tdc_config import *
try:
from pyroute2 import netns
from pyroute2 import IPRoute
netlink = True
except ImportError:
netlink = False
print("!!! Consider installing pyroute2 !!!")
class SubPlugin(TdcPlugin):
def __init__(self):
self.sub_class = 'ns/SubPlugin'
super().__init__()
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def pre_suite(self, testcount, testlist):
super().pre_suite(testcount, testlist)
def prepare_test(self, test):
if 'skip' in test and test['skip'] == 'yes':
return
if 'nsPlugin' not in test['plugins']:
return
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
if netlink == True:
self._nl_ns_create()
else:
self._ipr2_ns_create()
# Make sure the netns is visible in the fs
ticks = 20
while True:
if ticks == 0:
raise TimeoutError
self._proc_check()
try:
ns = self.args.NAMES['NS']
f = open('/run/netns/{}'.format(ns))
f.close()
break
except:
time.sleep(0.1)
ticks -= 1
continue
def pre_case(self, test, test_skip):
if self.args.verbose:
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
print('{}.pre_case'.format(self.sub_class))
if test_skip:
return
self.prepare_test(test)
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def post_case(self):
if self.args.verbose:
print('{}.post_case'.format(self.sub_class))
if netlink == True:
self._nl_ns_destroy()
else:
self._ipr2_ns_destroy()
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def post_suite(self, index):
if self.args.verbose:
print('{}.post_suite'.format(self.sub_class))
# Make sure we don't leak resources
cmd = self._replace_keywords("$IP -a netns del")
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
if self.args.verbose > 3:
print('_exec_cmd: command "{}"'.format(cmd))
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
subprocess.run(cmd, shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def adjust_command(self, stage, command):
super().adjust_command(stage, command)
cmdform = 'list'
cmdlist = list()
if self.args.verbose:
print('{}.adjust_command'.format(self.sub_class))
if not isinstance(command, list):
cmdform = 'str'
cmdlist = command.split()
else:
cmdlist = command
if stage == 'setup' or stage == 'execute' or stage == 'verify' or stage == 'teardown':
if self.args.verbose:
print('adjust_command: stage is {}; inserting netns stuff in command [{}] list [{}]'.format(stage, command, cmdlist))
cmdlist.insert(0, self.args.NAMES['NS'])
cmdlist.insert(0, 'exec')
cmdlist.insert(0, 'netns')
cmdlist.insert(0, self.args.NAMES['IP'])
else:
pass
if cmdform == 'str':
command = ' '.join(cmdlist)
else:
command = cmdlist
if self.args.verbose:
print('adjust_command: return command [{}]'.format(command))
return command
def _nl_ns_create(self):
ns = self.args.NAMES["NS"];
dev0 = self.args.NAMES["DEV0"];
dev1 = self.args.NAMES["DEV1"];
dummy = self.args.NAMES["DUMMY"];
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
if self.args.verbose:
print('{}._nl_ns_create'.format(self.sub_class))
netns.create(ns)
netns.pushns(newns=ns)
with IPRoute() as ip:
ip.link('add', ifname=dev1, kind='veth', peer={'ifname': dev0, 'net_ns_fd':'/proc/1/ns/net'})
ip.link('add', ifname=dummy, kind='dummy')
ticks = 20
while True:
if ticks == 0:
raise TimeoutError
try:
dev1_idx = ip.link_lookup(ifname=dev1)[0]
dummy_idx = ip.link_lookup(ifname=dummy)[0]
ip.link('set', index=dev1_idx, state='up')
ip.link('set', index=dummy_idx, state='up')
break
except:
time.sleep(0.1)
ticks -= 1
continue
netns.popns()
with IPRoute() as ip:
ticks = 20
while True:
if ticks == 0:
raise TimeoutError
try:
dev0_idx = ip.link_lookup(ifname=dev0)[0]
ip.link('set', index=dev0_idx, state='up')
break
except:
time.sleep(0.1)
ticks -= 1
continue
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def _ipr2_ns_create_cmds(self):
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
cmds = []
ns = self.args.NAMES['NS']
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
cmds.append(self._replace_keywords('netns add {}'.format(ns)))
cmds.append(self._replace_keywords('link add $DEV1 type veth peer name $DEV0'))
cmds.append(self._replace_keywords('link set $DEV1 netns {}'.format(ns)))
cmds.append(self._replace_keywords('link add $DUMMY type dummy'.format(ns)))
cmds.append(self._replace_keywords('link set $DUMMY netns {}'.format(ns)))
cmds.append(self._replace_keywords('netns exec {} $IP link set $DEV1 up'.format(ns)))
cmds.append(self._replace_keywords('netns exec {} $IP link set $DUMMY up'.format(ns)))
cmds.append(self._replace_keywords('link set $DEV0 up'.format(ns)))
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
if self.args.device:
cmds.append(self._replace_keywords('link set $DEV2 netns {}'.format(ns)))
cmds.append(self._replace_keywords('netns exec {} $IP link set $DEV2 up'.format(ns)))
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
return cmds
def _ipr2_ns_create(self):
'''
Create the network namespace in which the tests will be run and set up
the required network devices for it.
'''
self._exec_cmd_batched('pre', self._ipr2_ns_create_cmds())
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def _nl_ns_destroy(self):
ns = self.args.NAMES['NS']
netns.remove(ns)
def _ipr2_ns_destroy_cmd(self):
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
return self._replace_keywords('netns delete {}'.format(self.args.NAMES['NS']))
def _ipr2_ns_destroy(self):
'''
Destroy the network namespace for testing (and any associated network
devices as well)
'''
self._exec_cmd('post', self._ipr2_ns_destroy_cmd())
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
@cached_property
def _proc(self):
ip = self._replace_keywords("$IP -b -")
proc = subprocess.Popen(ip,
shell=True,
stdin=subprocess.PIPE,
env=ENVIR)
return proc
def _proc_check(self):
proc = self._proc
proc.poll()
if proc.returncode is not None and proc.returncode != 0:
raise RuntimeError("iproute2 exited with an error code")
def _exec_cmd(self, stage, command):
'''
Perform any required modifications on an executable command, then run
it in a subprocess and return the results.
'''
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
if self.args.verbose > 3:
print('_exec_cmd: command "{}"'.format(command))
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
proc = self._proc
proc.stdin.write((command + '\n').encode())
proc.stdin.flush()
if self.args.verbose > 3:
print('_exec_cmd proc: {}'.format(proc))
self._proc_check()
selftests/tc-testing: localize test resources As of today, the current tdc architecture creates one netns and uses it to run all tests. This assumption was embedded into the nsPlugin which carried over as how the tests were written. The tdc tests are by definition self contained and can, theoretically, run in parallel. Even though in the kernel they will serialize over the rtnl lock, we should expect a significant speedup of the total wall time for the entire test suite, which is hitting close to 1100 tests at this point. A first step to achieve this goal is to remove sharing of global resources like veth/dummy interfaces and the netns. In this patch we 'localize' these resources on a per test basis. Each test gets it's own netns, VETH/dummy interfaces. The resources are spawned in the pre_suite phase, where tdc will prepare all netns and interfaces for all tests. This is done in order to avoid concurrency issues with netns / interfaces spawning and commands using them. As tdc progresses, the resources are deleted after each test finishes executing. Tests that don't use the nsPlugin still run under the root namespace, but are now required to manage any external resources like interfaces. These cannot be parallelized as their definition doesn't allow it. On the other hand, when using the nsPlugin, tests don't need to create dummy/veth interfaces as these are handled already. Tested-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2023-09-19 10:54:01 -03:00
def _exec_cmd_batched(self, stage, commands):
for cmd in commands:
self._exec_cmd(stage, cmd)
def _replace_keywords(self, cmd):
"""
For a given executable command, substitute any known
variables contained within NAMES with the correct values
"""
tcmd = Template(cmd)
subcmd = tcmd.safe_substitute(self.args.NAMES)
return subcmd