ad9bca2e92
- Ability for pod to be Completed, unlike in Statefulset - eg. if 3 pods are running and first one finishes, all 3 must be running until all 3 are done. With this setup, the first finished pod can remain in Completed state. - Fixed shutdown order - crawler pods now correctly shutdown first before redis pods, by switching to background deletion. - Pod priority decreases with scale: 1st instance of a new crawl can preempt 3rd or 2nd instance of another crawl - Create priority classes upto 'max_crawl_scale, configured in values.yaml - Improved scale change reconciliation: if increasing scale, immediately scale up. If decreasing scale, graceful stop scaled-down instance to complete via redis 'stopone' key, wait until they exit with Completed state before adjust status.scale / removing scaled down pods. Ensures unaccepted interrupts don't cause scaled down data to be deleted. - Redis pod remains inactive until crawler is first active, or after no crawl pods are active for 60 seconds - Configurable Redis storage with 'redis_storage' value, set to 3Gi by default - CrawlJob deletion starts as soon as post-finish crawl operations are run - Post-crawl operations get their own redis instance, since one during response is being cleaned up in finalizer - Finalizer ignores request with incorrect state (returns 400 if reported as not finished while crawl is finished) - Current resource usage added to status - Profile browser: also manage single pod directly without statefulset for consistency. - Restart pods via restartTime value: if spec.restartTime != status.restartTime, clear out pods and update status.restartTime (using OnDelete policy to avoid recreate loops in edge cases). - Update to latest metacontroller (v4.11.0) - Add --restartOnError flag for crawler (for browsertrix-crawler 0.11.0) - Failed crawl logging: dd 'fail_crawl()' to be used for failing a crawl, which prints logs for default container (if enabled) as well as pod status - tests: check other finished states to avoid stuck in infinite loop if crawl fails - tests: disable disk utilization check, which adds unpredictability to crawl testing! fixes #1147 --------- Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
433 lines
14 KiB
Python
433 lines
14 KiB
Python
""" shared crawl manager implementation """
|
|
|
|
import os
|
|
import asyncio
|
|
import secrets
|
|
import json
|
|
import base64
|
|
|
|
from datetime import timedelta
|
|
|
|
from .k8sapi import K8sAPI
|
|
from .models import S3Storage
|
|
from .utils import dt_now, to_k8s_date
|
|
|
|
|
|
# ============================================================================
|
|
class CrawlManager(K8sAPI):
|
|
"""abstract crawl manager"""
|
|
|
|
def __init__(self):
|
|
super().__init__()
|
|
self.job_image = os.environ["JOB_IMAGE"]
|
|
self.job_image_pull_policy = os.environ.get("JOB_PULL_POLICY", "Always")
|
|
|
|
self.cron_namespace = os.environ.get("CRON_NAMESPACE", "default")
|
|
|
|
self._default_storages = {}
|
|
|
|
self.loop = asyncio.get_running_loop()
|
|
|
|
# pylint: disable=too-many-arguments
|
|
async def run_profile_browser(
|
|
self,
|
|
userid,
|
|
oid,
|
|
url,
|
|
storage=None,
|
|
storage_name=None,
|
|
baseprofile=None,
|
|
profile_path=None,
|
|
):
|
|
"""run browser for profile creation"""
|
|
|
|
# if default storage, use name and path + profiles/
|
|
if storage:
|
|
storage_name = storage.name
|
|
storage_path = storage.path + "profiles/"
|
|
# otherwise, use storage name and existing path from secret
|
|
else:
|
|
storage_path = ""
|
|
|
|
await self.check_storage(storage_name)
|
|
|
|
browserid = f"prf-{secrets.token_hex(5)}"
|
|
|
|
params = {
|
|
"id": browserid,
|
|
"userid": str(userid),
|
|
"oid": str(oid),
|
|
"storage_name": storage_name,
|
|
"storage_path": storage_path or "",
|
|
"base_profile": baseprofile or "",
|
|
"profile_filename": profile_path,
|
|
"idle_timeout": os.environ.get("IDLE_TIMEOUT", "60"),
|
|
"url": url,
|
|
"vnc_password": secrets.token_hex(16),
|
|
"expire_time": to_k8s_date(dt_now() + timedelta(seconds=30)),
|
|
}
|
|
|
|
data = self.templates.env.get_template("profile_job.yaml").render(params)
|
|
|
|
await self.create_from_yaml(data)
|
|
|
|
return browserid
|
|
|
|
async def add_crawl_config(
|
|
self,
|
|
crawlconfig,
|
|
storage,
|
|
run_now,
|
|
out_filename,
|
|
profile_filename,
|
|
):
|
|
"""add new crawl, store crawl config in configmap"""
|
|
|
|
if storage.type == "default":
|
|
storage_name = storage.name
|
|
storage_path = storage.path
|
|
else:
|
|
storage_name = str(crawlconfig.oid)
|
|
storage_path = ""
|
|
|
|
await self.check_storage(storage_name)
|
|
|
|
# Create Config Map
|
|
await self._create_config_map(
|
|
crawlconfig,
|
|
USER_ID=str(crawlconfig.modifiedBy),
|
|
ORG_ID=str(crawlconfig.oid),
|
|
CRAWL_CONFIG_ID=str(crawlconfig.id),
|
|
STORE_PATH=storage_path,
|
|
STORE_FILENAME=out_filename,
|
|
STORAGE_NAME=storage_name,
|
|
PROFILE_FILENAME=profile_filename,
|
|
INITIAL_SCALE=str(crawlconfig.scale),
|
|
CRAWL_TIMEOUT=str(crawlconfig.crawlTimeout or 0),
|
|
MAX_CRAWL_SIZE=str(crawlconfig.maxCrawlSize or 0)
|
|
# REV=str(crawlconfig.rev),
|
|
)
|
|
|
|
crawl_id = None
|
|
|
|
if run_now:
|
|
crawl_id = await self.create_crawl_job(
|
|
crawlconfig, str(crawlconfig.modifiedBy)
|
|
)
|
|
|
|
await self._update_scheduled_job(crawlconfig, crawlconfig.schedule)
|
|
|
|
return crawl_id
|
|
|
|
async def create_crawl_job(self, crawlconfig, userid: str):
|
|
"""create new crawl job from config"""
|
|
cid = str(crawlconfig.id)
|
|
|
|
return await self.new_crawl_job(
|
|
cid,
|
|
userid,
|
|
crawlconfig.oid,
|
|
crawlconfig.scale,
|
|
crawlconfig.crawlTimeout,
|
|
crawlconfig.maxCrawlSize,
|
|
manual=True,
|
|
)
|
|
|
|
async def update_crawl_config(self, crawlconfig, update, profile_filename=None):
|
|
"""Update the schedule or scale for existing crawl config"""
|
|
|
|
has_sched_update = update.schedule is not None
|
|
has_scale_update = update.scale is not None
|
|
has_timeout_update = update.crawlTimeout is not None
|
|
has_max_crawl_size_update = update.maxCrawlSize is not None
|
|
has_config_update = update.config is not None
|
|
|
|
if has_sched_update:
|
|
await self._update_scheduled_job(crawlconfig, update.schedule)
|
|
|
|
if (
|
|
has_scale_update
|
|
or has_config_update
|
|
or has_timeout_update
|
|
or profile_filename
|
|
or has_max_crawl_size_update
|
|
):
|
|
await self._update_config_map(
|
|
crawlconfig,
|
|
update,
|
|
profile_filename,
|
|
has_config_update,
|
|
)
|
|
|
|
return True
|
|
|
|
# pylint: disable=unused-argument
|
|
async def check_storage(self, storage_name, is_default=False):
|
|
"""Check if storage is valid by trying to get the storage secret
|
|
Will throw if not valid, otherwise return True"""
|
|
await self._get_storage_secret(storage_name)
|
|
return True
|
|
|
|
async def update_org_storage(self, oid, userid, storage):
|
|
"""Update storage by either creating a per-org secret, if using custom storage
|
|
or deleting per-org secret, if using default storage"""
|
|
org_storage_name = f"storage-{oid}"
|
|
if storage.type == "default":
|
|
try:
|
|
await self.core_api.delete_namespaced_secret(
|
|
org_storage_name,
|
|
namespace=self.namespace,
|
|
propagation_policy="Foreground",
|
|
)
|
|
# pylint: disable=bare-except
|
|
except:
|
|
pass
|
|
|
|
return
|
|
|
|
labels = {"btrix.org": oid, "btrix.user": userid}
|
|
|
|
crawl_secret = self.client.V1Secret(
|
|
metadata={
|
|
"name": org_storage_name,
|
|
"namespace": self.namespace,
|
|
"labels": labels,
|
|
},
|
|
string_data={
|
|
"STORE_ENDPOINT_URL": storage.endpoint_url,
|
|
"STORE_ACCESS_KEY": storage.access_key,
|
|
"STORE_SECRET_KEY": storage.secret_key,
|
|
},
|
|
)
|
|
|
|
try:
|
|
await self.core_api.create_namespaced_secret(
|
|
namespace=self.namespace, body=crawl_secret
|
|
)
|
|
|
|
# pylint: disable=bare-except
|
|
except:
|
|
await self.core_api.patch_namespaced_secret(
|
|
name=org_storage_name, namespace=self.namespace, body=crawl_secret
|
|
)
|
|
|
|
async def get_default_storage_access_endpoint(self, name):
|
|
"""Get access_endpoint for default storage"""
|
|
return (await self.get_default_storage(name)).access_endpoint_url
|
|
|
|
async def get_default_storage(self, name):
|
|
"""get default storage"""
|
|
if name not in self._default_storages:
|
|
storage_secret = await self._get_storage_secret(name)
|
|
|
|
access_endpoint_url = self._secret_data(
|
|
storage_secret, "STORE_ACCESS_ENDPOINT_URL"
|
|
)
|
|
endpoint_url = self._secret_data(storage_secret, "STORE_ENDPOINT_URL")
|
|
access_key = self._secret_data(storage_secret, "STORE_ACCESS_KEY")
|
|
secret_key = self._secret_data(storage_secret, "STORE_SECRET_KEY")
|
|
region = self._secret_data(storage_secret, "STORE_REGION") or ""
|
|
use_access_for_presign = (
|
|
self._secret_data(storage_secret, "STORE_USE_ACCESS_FOR_PRESIGN") == "1"
|
|
)
|
|
|
|
self._default_storages[name] = S3Storage(
|
|
access_key=access_key,
|
|
secret_key=secret_key,
|
|
endpoint_url=endpoint_url,
|
|
access_endpoint_url=access_endpoint_url,
|
|
region=region,
|
|
use_access_for_presign=use_access_for_presign,
|
|
)
|
|
|
|
return self._default_storages[name]
|
|
|
|
async def get_profile_browser_metadata(self, browserid):
|
|
"""get browser profile labels"""
|
|
try:
|
|
browser = await self.get_profile_browser(browserid)
|
|
|
|
# pylint: disable=bare-except
|
|
except:
|
|
return {}
|
|
|
|
return browser["metadata"]["labels"]
|
|
|
|
async def get_configmap(self, cid):
|
|
"""get configmap by id"""
|
|
return await self.core_api.read_namespaced_config_map(
|
|
name=f"crawl-config-{cid}", namespace=self.namespace
|
|
)
|
|
|
|
async def ping_profile_browser(self, browserid):
|
|
"""return ping profile browser"""
|
|
expire_at = dt_now() + timedelta(seconds=30)
|
|
await self._patch_job(
|
|
browserid, {"expireTime": to_k8s_date(expire_at)}, "profilejobs"
|
|
)
|
|
|
|
async def rollover_restart_crawl(self, crawl_id, oid):
|
|
"""Rolling restart of crawl by updating restartTime field"""
|
|
update = to_k8s_date(dt_now())
|
|
return await self._patch_job(crawl_id, {"restartTime": update})
|
|
|
|
async def scale_crawl(self, crawl_id, oid, scale=1):
|
|
"""Set the crawl scale (job parallelism) on the specified job"""
|
|
return await self._patch_job(crawl_id, {"scale": scale})
|
|
|
|
async def shutdown_crawl(self, crawl_id, oid, graceful=True):
|
|
"""Request a crawl cancelation or stop by calling an API
|
|
on the job pod/container, returning the result"""
|
|
if graceful:
|
|
patch = {"stopping": True}
|
|
return await self._patch_job(crawl_id, patch)
|
|
|
|
return await self.delete_crawl_job(crawl_id)
|
|
|
|
async def delete_crawl_configs_for_org(self, org):
|
|
"""Delete all crawl configs for given org"""
|
|
return await self._delete_crawl_configs(f"btrix.org={org}")
|
|
|
|
async def delete_crawl_config_by_id(self, cid):
|
|
"""Delete all crawl configs by id"""
|
|
return await self._delete_crawl_configs(f"btrix.crawlconfig={cid}")
|
|
|
|
# ========================================================================
|
|
# Internal Methods
|
|
def _secret_data(self, secret, name):
|
|
"""decode secret data"""
|
|
return base64.standard_b64decode(secret.data[name]).decode()
|
|
|
|
async def _create_config_map(self, crawlconfig, **data):
|
|
"""Create Config Map based on CrawlConfig"""
|
|
data["crawl-config.json"] = json.dumps(crawlconfig.get_raw_config())
|
|
|
|
labels = {
|
|
"btrix.crawlconfig": str(crawlconfig.id),
|
|
"btrix.org": str(crawlconfig.oid),
|
|
}
|
|
|
|
config_map = self.client.V1ConfigMap(
|
|
metadata={
|
|
"name": f"crawl-config-{crawlconfig.id}",
|
|
"namespace": self.namespace,
|
|
"labels": labels,
|
|
},
|
|
data=data,
|
|
)
|
|
|
|
return await self.core_api.create_namespaced_config_map(
|
|
namespace=self.namespace, body=config_map
|
|
)
|
|
|
|
async def _get_storage_secret(self, storage_name):
|
|
"""Check if storage_name is valid by checking existing secret"""
|
|
try:
|
|
return await self.core_api.read_namespaced_secret(
|
|
f"storage-{storage_name}",
|
|
namespace=self.namespace,
|
|
)
|
|
# pylint: disable=broad-except
|
|
except Exception:
|
|
# pylint: disable=broad-exception-raised,raise-missing-from
|
|
raise Exception(f"Storage {storage_name} not found")
|
|
|
|
return None
|
|
|
|
async def _delete_crawl_configs(self, label):
|
|
"""Delete Crawl Cron Job and all dependent resources, including configmap and secrets"""
|
|
|
|
await self.batch_api.delete_collection_namespaced_cron_job(
|
|
namespace=self.cron_namespace,
|
|
label_selector=label,
|
|
propagation_policy="Foreground",
|
|
)
|
|
|
|
await self.core_api.delete_collection_namespaced_config_map(
|
|
namespace=self.namespace,
|
|
label_selector=label,
|
|
propagation_policy="Foreground",
|
|
)
|
|
|
|
async def _update_scheduled_job(self, crawlconfig, schedule):
|
|
"""create or remove cron job based on crawlconfig schedule"""
|
|
cid = str(crawlconfig.id)
|
|
|
|
cron_job_id = f"sched-{cid[:12]}"
|
|
cron_job = None
|
|
try:
|
|
cron_job = await self.batch_api.read_namespaced_cron_job(
|
|
name=cron_job_id,
|
|
namespace=self.cron_namespace,
|
|
)
|
|
# pylint: disable=bare-except
|
|
except:
|
|
pass
|
|
|
|
# if no schedule, delete cron_job if exists and we're done
|
|
if not crawlconfig.schedule:
|
|
if cron_job:
|
|
await self.batch_api.delete_namespaced_cron_job(
|
|
name=cron_job.metadata.name, namespace=self.cron_namespace
|
|
)
|
|
return
|
|
|
|
# if cron job exists, just patch schedule
|
|
if cron_job:
|
|
if crawlconfig.schedule != cron_job.spec.schedule:
|
|
cron_job.spec.schedule = crawlconfig.schedule
|
|
|
|
await self.batch_api.patch_namespaced_cron_job(
|
|
name=cron_job.metadata.name,
|
|
namespace=self.cron_namespace,
|
|
body=cron_job,
|
|
)
|
|
return
|
|
|
|
params = {
|
|
"id": cron_job_id,
|
|
"cid": str(crawlconfig.id),
|
|
"image": self.job_image,
|
|
"image_pull_policy": self.job_image_pull_policy,
|
|
"schedule": schedule,
|
|
}
|
|
|
|
data = self.templates.env.get_template("crawl_cron_job.yaml").render(params)
|
|
|
|
await self.create_from_yaml(data, self.cron_namespace)
|
|
|
|
return cron_job_id
|
|
|
|
async def _update_config_map(
|
|
self,
|
|
crawlconfig,
|
|
update,
|
|
profile_filename=None,
|
|
update_config=False,
|
|
):
|
|
config_map = await self.get_configmap(crawlconfig.id)
|
|
|
|
if update.scale is not None:
|
|
config_map.data["INITIAL_SCALE"] = str(update.scale)
|
|
|
|
if update.crawlTimeout is not None:
|
|
config_map.data["CRAWL_TIMEOUT"] = str(update.crawlTimeout)
|
|
|
|
if update.maxCrawlSize is not None:
|
|
config_map.data["MAX_CRAWL_SIZE"] = str(update.maxCrawlSize)
|
|
|
|
if update.crawlFilenameTemplate is not None:
|
|
config_map.data["STORE_FILENAME"] = update.crawlFilenameTemplate
|
|
|
|
if profile_filename is not None:
|
|
config_map.data["PROFILE_FILENAME"] = profile_filename
|
|
|
|
if update_config:
|
|
config_map.data["crawl-config.json"] = json.dumps(
|
|
crawlconfig.get_raw_config()
|
|
)
|
|
|
|
await self.core_api.patch_namespaced_config_map(
|
|
name=config_map.metadata.name, namespace=self.namespace, body=config_map
|
|
)
|