00eb62214d
* basecrawl refactor: make crawls db more generic, supporting different types of 'base crawls': crawls, uploads, manual archives - move shared functionality to basecrawl.py - create a base BaseCrawl object, which contains start / finish time, metadata and files array - create BaseCrawlOps, base class for CrawlOps, which supports base crawl deletion, querying and collection add/remove * uploads api: (part of #929) - new UploadCrawl object which extends BaseCrawl, has name and description - support multipart form data data upload to /uploads/formdata - support streaming upload of a single file via /uploads/stream, using botocore multipart upload to upload to s3-endpoint in parts - require 'filename' param to set upload filename for streaming uploads (otherwise use form data names) - sanitize filename, place uploads in /uploads/<uuid>/<sanitized-filename>-<random>.wacz - uploads have internal id 'upload-<uuid>' - create UploadedCrawl object with CrawlFiles pointing to the newly uploaded files, set state to 'complete' - handle upload failures, abort multipart upload - ensure uploads added within org bucket path - return id / added when adding new UploadedCrawl - support listing, deleting, and patch /uploads - support upload details via /replay.json to support for replay - add support for 'replaceId=<id>', which would remove all previous files in upload after new upload succeeds. if replaceId doesn't exist, create new upload. (only for stream endpoint so far). - support patching upload metadata: notes, tags and name on uploads (UpdateUpload extends UpdateCrawl and adds 'name') * base crawls api: Add /all-crawls list and delete endpoints for all crawl types (without resources) - support all-crawls/<id>/replay.json with resources - Use ListCrawlOut model for /all-crawls list endpoint - Extend BaseCrawlOut from ListCrawlOut, add type - use 'type: crawl' for crawls and 'type: upload' for uploads - migration: ensure all previous crawl objects / missing type are set to 'type: crawl' - indexes: add db indices on 'type' field and with 'type' field and oid, cid, finished, state * tests: add test for multipart and streaming upload, listing uploads, deleting upload - add sample WACZ for upload testing: 'example.wacz' and 'example-2.wacz' * collections: support adding and remove both crawls and uploads via base crawl - include collection_ids in /all-crawls list - collections replay.json can include both crawls and uploads bump version to 1.6.0-beta.2 --------- Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
72 lines
2.6 KiB
Python
72 lines
2.6 KiB
Python
"""
|
|
Migration 0005 - Updating scheduled cron jobs after Operator changes
|
|
"""
|
|
from btrixcloud.crawlconfigs import CrawlConfig, UpdateCrawlConfig
|
|
from btrixcloud.crawlmanager import CrawlManager
|
|
from btrixcloud.migrations import BaseMigration
|
|
|
|
|
|
MIGRATION_VERSION = "0005"
|
|
|
|
|
|
class Migration(BaseMigration):
|
|
"""Migration class."""
|
|
|
|
def __init__(self, mdb, migration_version=MIGRATION_VERSION):
|
|
super().__init__(mdb, migration_version)
|
|
|
|
async def migrate_up(self):
|
|
"""Perform migration up.
|
|
|
|
Find existing workflows with schedule and create new crawl_cron_jobs
|
|
from template for each, then delete existing scheduled jobs from
|
|
crawler namespace.
|
|
|
|
Additionally update the configmap for crawl configs with scale > 1
|
|
or crawlTimeout > 0.
|
|
"""
|
|
# pylint: disable=too-many-locals, duplicate-code
|
|
crawl_configs = self.mdb["crawl_configs"]
|
|
crawl_manager = CrawlManager()
|
|
|
|
# Update configmap for crawl configs that have non-zero timeout or scale > 1
|
|
match_query = {
|
|
"$or": [
|
|
{"crawlTimeout": {"$gt": 0}},
|
|
{"scale": {"$gt": 1}},
|
|
{"schedule": {"$nin": ["", None]}},
|
|
]
|
|
}
|
|
configs_to_update = [res async for res in crawl_configs.find(match_query)]
|
|
for config_dict in configs_to_update:
|
|
config = CrawlConfig.from_dict(config_dict)
|
|
print(
|
|
f"Updating Crawl Config {config.id}: schedule: {config.schedule}, "
|
|
+ f"timeout: {config.crawlTimeout}, scale: {config.scale}"
|
|
)
|
|
try:
|
|
await crawl_manager.update_crawl_config(
|
|
config,
|
|
UpdateCrawlConfig(
|
|
scale=config.scale,
|
|
crawlTimeout=config.crawlTimeout,
|
|
schedule=config.schedule,
|
|
),
|
|
)
|
|
# pylint: disable=broad-except
|
|
except Exception as exc:
|
|
print(
|
|
"Skip crawl config migration due to error, likely missing config",
|
|
exc,
|
|
)
|
|
|
|
# Delete existing scheduled jobs from crawler namespace
|
|
print("Deleting cronjobs from crawler namespace")
|
|
await crawl_manager.batch_api.delete_collection_namespaced_cron_job(
|
|
namespace=crawl_manager.namespace
|
|
)
|
|
result = await crawl_manager.batch_api.list_namespaced_cron_job(
|
|
namespace=crawl_manager.namespace
|
|
)
|
|
assert len(result.items) == 0
|