[BUG] GremlinServerError with Complex Query

[EricaGuoQiuyu]

Describe the bug
The deadline was exceeded on the complex queries with more than 10M-scale results, while working well on queries with less than million-scale results.

To Reproduce
Steps to reproduce the behavior:

1. Download the LSQB dataset sf3
2. Download the LSQB loader load_lsqb which reused your ldbc loader via this aliyundrive link
3. Run :

import graphscope as gs
import pandas as pd
import os
import time
gs.set_option(vineyard_shared_mem="400G")

sess = gs.session(cluster_type='hosts')

from load_lsqb import load_lsqb
graph = load_lsqb(sess, prefix="/root/social-network-sf3-projected-fk/", directed=True) #modify prefix
print(graph.schema)

interactive = gs.gremlin(graph)

#modify the Q1.gremlin prefix
#Q1 has about 100M-scale results on sf3
'''
Q1.gremlin:
g.V().match(__.as('liker').out('likes').as('Message'),
__.as('Message').out('hasTag').as('Tag'),
__.as('Message').out('hasCreator').as('creator'),
__.as('Message').in('replyOf').as('Comment'))
'''

with open('/root/Q1.gremlin','r') as f:
  gremlin_query = f.read()

print("start querying")
start = time.time()
result = interactive.execute(gremlin_query).all()
end = time.time()

print("count: ", result)
print("time: ", end-start)

sess.close()

Expected behavior

GremlinServerError: 500: Status{code=DEADLINE_EXCEEDED, description=deadline exceeded after 599.999663051s. [closed=[], 
committed=[remote_addr=127.0.0.1/127.0.0.1:8234]], cause=null}

Environment (please complete the following information):

• GraphScope version: [ v0.22 ]
• OS: [ Linux ]
• Version [ ubuntu20.04 ]

Additional context
Everything works well on the queries with less than million-scale results like Q2.gremlin:

g.V().match(__.as('person1').both('knows').as('person2'),
__.as('person1').both('knows').as('person3'),
__.as('person2').both('knows').as('person3'),
__.as('person1').out('isLocatedIn').as('city1'),
__.as('city1').out('isPartOf').as('country'),
__.as('person2').out('isLocatedIn').as('city2'),
__.as('city2').out('isPartOf').as('country'),
__.as('person3').out('isLocatedIn').as('city3'),
__.as('city3').out('isPartOf').as('country')).count()

[longbinlai]

Thank you for posting the issue. We are doing an internal test using LSQB, and hopefully we can figure out what the problem really is. Get back to you soon.