diff options
Diffstat (limited to '')
| -rw-r--r-- | taskcluster/gecko_taskgraph/optimize/bugbug.py | 323 |
1 files changed, 323 insertions, 0 deletions
diff --git a/taskcluster/gecko_taskgraph/optimize/bugbug.py b/taskcluster/gecko_taskgraph/optimize/bugbug.py new file mode 100644 index 0000000000..1ae717b53c --- /dev/null +++ b/taskcluster/gecko_taskgraph/optimize/bugbug.py @@ -0,0 +1,323 @@ +# This Source Code Form is subject to the terms of the Mozilla Public +# License, v. 2.0. If a copy of the MPL was not distributed with this +# file, You can obtain one at http://mozilla.org/MPL/2.0/. + + +from collections import defaultdict +from fnmatch import fnmatch +from urllib.parse import urlsplit + +from taskgraph.optimize.base import OptimizationStrategy, register_strategy, registry + +from gecko_taskgraph.util.bugbug import ( + CT_HIGH, + CT_LOW, + CT_MEDIUM, + BugbugTimeoutException, + push_schedules, +) +from gecko_taskgraph.util.hg import get_push_data + +FALLBACK = "skip-unless-has-relevant-tests" + + +def merge_bugbug_replies(data, new_data): + """Merge a bugbug reply (stored in the `new_data` argument) into another (stored + in the `data` argument). + """ + for key, value in new_data.items(): + if isinstance(value, dict): + if key not in data: + data[key] = {} + + if len(value) == 0: + continue + + dict_value = next(iter(value.values())) + if isinstance(dict_value, list): + for name, configs in value.items(): + if name not in data[key]: + data[key][name] = set() + + data[key][name].update(configs) + else: + for name, confidence in value.items(): + if name not in data[key] or data[key][name] < confidence: + data[key][name] = confidence + elif isinstance(value, list): + if key not in data: + data[key] = set() + + data[key].update(value) + + +@register_strategy("bugbug-low", args=(CT_LOW,)) +@register_strategy("bugbug-medium", args=(CT_MEDIUM,)) +@register_strategy("bugbug-high", args=(CT_HIGH,)) +@register_strategy("bugbug-tasks-medium", args=(CT_MEDIUM, True)) +@register_strategy("bugbug-tasks-high", args=(CT_HIGH, True)) +@register_strategy("bugbug-reduced", args=(CT_MEDIUM, True, True)) +@register_strategy("bugbug-reduced-fallback", args=(CT_MEDIUM, True, True, FALLBACK)) +@register_strategy("bugbug-reduced-high", args=(CT_HIGH, True, True)) +@register_strategy("bugbug-reduced-manifests", args=(CT_MEDIUM, False, True)) +@register_strategy( + "bugbug-reduced-manifests-config-selection-low", + args=(CT_LOW, False, True, None, 1, True), +) +@register_strategy( + "bugbug-reduced-manifests-config-selection", + args=(CT_MEDIUM, False, True, None, 1, True), +) +@register_strategy( + "bugbug-reduced-manifests-fallback-low", args=(CT_LOW, False, True, FALLBACK) +) +@register_strategy( + "bugbug-reduced-manifests-fallback", args=(CT_MEDIUM, False, True, FALLBACK) +) +@register_strategy( + "bugbug-reduced-manifests-fallback-last-10-pushes", + args=(0.3, False, True, FALLBACK, 10), +) +class BugBugPushSchedules(OptimizationStrategy): + """Query the 'bugbug' service to retrieve relevant tasks and manifests. + + Args: + confidence_threshold (float): The minimum confidence threshold (in + range [0, 1]) needed for a task to be scheduled. + tasks_only (bool): Whether or not to only use tasks and no groups + (default: False) + use_reduced_tasks (bool): Whether or not to use the reduced set of tasks + provided by the bugbug service (default: False). + fallback (str): The fallback strategy to use if there + was a failure in bugbug (default: None) + num_pushes (int): The number of pushes to consider for the selection + (default: 1). + select_configs (bool): Whether to select configurations for manifests + too (default: False). + """ + + def __init__( + self, + confidence_threshold, + tasks_only=False, + use_reduced_tasks=False, + fallback=None, + num_pushes=1, + select_configs=False, + ): + self.confidence_threshold = confidence_threshold + self.use_reduced_tasks = use_reduced_tasks + self.fallback = fallback + self.tasks_only = tasks_only + self.num_pushes = num_pushes + self.select_configs = select_configs + self.timedout = False + + def should_remove_task(self, task, params, importance): + project = params["project"] + + if project not in ("autoland", "try"): + return False + + current_push_id = int(params["pushlog_id"]) + + branch = urlsplit(params["head_repository"]).path.strip("/") + rev = params["head_rev"] + + if self.timedout: + return registry[self.fallback].should_remove_task(task, params, importance) + + data = {} + + start_push_id = current_push_id - self.num_pushes + 1 + if self.num_pushes != 1: + push_data = get_push_data( + params["head_repository"], project, start_push_id, current_push_id - 1 + ) + + for push_id in range(start_push_id, current_push_id + 1): + if push_id == current_push_id: + rev = params["head_rev"] + else: + rev = push_data[push_id]["changesets"][-1] + + try: + new_data = push_schedules(branch, rev) + merge_bugbug_replies(data, new_data) + except BugbugTimeoutException: + if not self.fallback: + raise + + self.timedout = True + return self.should_remove_task(task, params, importance) + + key = "reduced_tasks" if self.use_reduced_tasks else "tasks" + tasks = { + task + for task, confidence in data.get(key, {}).items() + if confidence >= self.confidence_threshold + } + + test_manifests = task.attributes.get("test_manifests") + if test_manifests is None or self.tasks_only: + if data.get("known_tasks") and task.label not in data["known_tasks"]: + return False + + if task.label not in tasks: + return True + + return False + + # If a task contains more than one group, use the max confidence. + groups = data.get("groups", {}) + confidences = [c for g, c in groups.items() if g in test_manifests] + if not confidences or max(confidences) < self.confidence_threshold: + return True + + # If the task configuration doesn't match the ones selected by bugbug for + # the manifests, optimize out. + if self.select_configs: + selected_groups = [ + g + for g, c in groups.items() + if g in test_manifests and c > self.confidence_threshold + ] + + config_groups = data.get("config_groups", defaultdict(list)) + + # Configurations returned by bugbug are in a format such as + # `test-windows10-64/opt-*-e10s`, while task labels are like + # test-windows10-64-qr/opt-mochitest-browser-chrome-e10s-6. + # In order to match the strings, we need to ignore the chunk number + # from the task label. + parts = task.label.split("-") + label_without_chunk_number = "-".join( + parts[:-1] if parts[-1].isdigit() else parts + ) + + if not any( + fnmatch(label_without_chunk_number, config) + for group in selected_groups + for config in config_groups[group] + ): + return True + + # Store group importance so future optimizers can access it. + for manifest in test_manifests: + if manifest not in groups: + continue + + confidence = groups[manifest] + if confidence >= CT_HIGH: + importance[manifest] = "high" + elif confidence >= CT_MEDIUM: + importance[manifest] = "medium" + elif confidence >= CT_LOW: + importance[manifest] = "low" + else: + importance[manifest] = "lowest" + + return False + + +@register_strategy("platform-debug") +class SkipUnlessDebug(OptimizationStrategy): + """Only run debug platforms.""" + + def should_remove_task(self, task, params, arg): + return ( + "build_type" in task.attributes and task.attributes["build_type"] != "debug" + ) + + +@register_strategy("platform-disperse") +@register_strategy("platform-disperse-no-unseen", args=(None, 0)) +@register_strategy( + "platform-disperse-only-one", + args=( + { + "high": 1, + "medium": 1, + "low": 1, + "lowest": 0, + }, + 0, + ), +) +class DisperseGroups(OptimizationStrategy): + """Disperse groups across test configs. + + Each task has an associated 'importance' dict passed in via the arg. This + is of the form `{<group>: <importance>}`. + + Where 'group' is a test group id (usually a path to a manifest), and 'importance' is + one of `{'lowest', 'low', 'medium', 'high'}`. + + Each importance value has an associated 'count' as defined in + `self.target_counts`. It guarantees that 'manifest' will run in at least + 'count' different configurations (assuming there are enough tasks + containing 'manifest'). + + On configurations that haven't been seen before, we'll increase the target + count by `self.unseen_modifier` to increase the likelihood of scheduling a + task on that configuration. + + Args: + target_counts (dict): Override DEFAULT_TARGET_COUNTS with custom counts. This + is a dict mapping the importance value ('lowest', 'low', etc) to the + minimum number of configurations manifests with this value should run + on. + + unseen_modifier (int): Override DEFAULT_UNSEEN_MODIFIER to a custom + value. This is the amount we'll increase 'target_count' by for unseen + configurations. + """ + + DEFAULT_TARGET_COUNTS = { + "high": 3, + "medium": 2, + "low": 1, + "lowest": 0, + } + DEFAULT_UNSEEN_MODIFIER = 1 + + def __init__(self, target_counts=None, unseen_modifier=DEFAULT_UNSEEN_MODIFIER): + self.target_counts = self.DEFAULT_TARGET_COUNTS.copy() + if target_counts: + self.target_counts.update(target_counts) + self.unseen_modifier = unseen_modifier + + self.count = defaultdict(int) + self.seen_configurations = set() + + def should_remove_task(self, task, params, importance): + test_manifests = task.attributes.get("test_manifests") + test_platform = task.attributes.get("test_platform") + + if not importance or not test_manifests or not test_platform: + return False + + # Build the test configuration key. + key = test_platform + if "unittest_variant" in task.attributes: + key += "-" + task.attributes["unittest_variant"] + + important_manifests = set(test_manifests) & set(importance) + for manifest in important_manifests: + target_count = self.target_counts[importance[manifest]] + + # If this configuration hasn't been seen before, increase the + # likelihood of scheduling the task. + if key not in self.seen_configurations: + target_count += self.unseen_modifier + + if self.count[manifest] < target_count: + # Update manifest counts and seen configurations. + self.seen_configurations.add(key) + for manifest in important_manifests: + self.count[manifest] += 1 + return False + + # Should remove task because all manifests have reached their + # importance count (or there were no important manifests). + return True |