re-claim sites after 1 hour instead of 2 so that sites don't have to wait as long to be brozzled again in case of kill -9 brozzler-worker

This commit is contained in:
Noah Levitt 2017-05-01 12:59:44 -07:00
parent 000d40c4dc
commit 52433ade78
3 changed files with 39 additions and 3 deletions

View file

@ -698,3 +698,39 @@ def test_honor_stop_request():
with pytest.raises(brozzler.CrawlStopped):
frontier.honor_stop_request(site)
def test_claim_site():
rr = doublethink.Rethinker('localhost', db='ignoreme')
frontier = brozzler.RethinkDbFrontier(rr)
rr.table('sites').delete().run() # clean slate
with pytest.raises(brozzler.NothingToClaim):
claimed_site = frontier.claim_site(worker_id='test_claim_site')
site = brozzler.Site(rr, {'seed': 'http://example.org/'})
brozzler.new_site(frontier, site)
claimed_site = frontier.claim_site(worker_id='test_claim_site')
assert claimed_site.id == site.id
assert claimed_site.claimed
assert claimed_site.last_claimed >= doublethink.utcnow() - datetime.timedelta(minutes=1)
with pytest.raises(brozzler.NothingToClaim):
claimed_site = frontier.claim_site(worker_id='test_claim_site')
# site last_claimed less than 1 hour ago still not to be reclaimed
claimed_site.last_claimed = doublethink.utcnow() - datetime.timedelta(minutes=55)
claimed_site.save()
with pytest.raises(brozzler.NothingToClaim):
claimed_site = frontier.claim_site(worker_id='test_claim_site')
# site last_claimed more than 1 hour ago can be reclaimed
site = claimed_site
claimed_site = None
site.last_claimed = doublethink.utcnow() - datetime.timedelta(minutes=65)
site.save()
claimed_site = frontier.claim_site(worker_id='test_claim_site')
assert claimed_site.id == site.id
# clean up
rr.table('sites').get(claimed_site.id).delete().run()