memory usage increase significantly caused by eth.contract()

[mwrrlvkixj]

• Version: 6.10.0
• Python: 3.9
• OS: osx
• pip freeze output

aiohttp==3.8.5
aiosignal==1.3.1
async-timeout==4.0.2
attrs==23.1.0
base58==2.1.1
bech32==1.2.0
bitarray==2.8.0
boto3==1.21.19
botocore==1.24.46
cachetools==4.2.4
certifi==2023.7.22
charset-normalizer==2.0.12
contourpy==1.1.1
cycler==0.12.1
cytoolz==0.12.2
datadog==0.46.0
eth-abi==4.2.1
eth-account==0.9.0
eth-hash==0.5.2
eth-keyfile==0.6.1
eth-keys==0.4.0
eth-rlp==0.3.0
eth-typing==3.5.1
eth-utils==2.2.1
fonttools==4.43.1
frozenlist==1.4.0
guppy3==3.1.4.post1
hexbytes==0.2.2
idna==3.4
importlib-resources==6.1.0
ipfshttpclient==0.8.0a2
Jinja2==3.1.2
jmespath==0.10.0
jsonschema==4.18.4
jsonschema-specifications==2023.7.1
kiwisolver==1.4.5
lru-dict==1.2.0
markdown-it-py==3.0.0
MarkupSafe==2.1.3
matplotlib==3.8.0
mdurl==0.1.2
memory-profiler==0.61.0
memray==1.10.0
more-itertools==9.1.0
multiaddr==0.0.9
multidict==6.0.4
netaddr==0.8.0
numpy==1.22.0
packaging==23.1
pandas==1.5.3
parsimonious==0.9.0
Pillow==10.1.0
pluggy==0.13.1
protobuf==4.24.4
psutil==5.9.6
psycopg2==2.9.6
psycopg2-binary==2.9.5
py==1.11.0
pycryptodome==3.18.0
Pygments==2.16.1
pyparsing==3.1.1
pytest==5.4.1
python-dateutil==2.8.2
pytz==2021.3
pyunormalize==15.0.0
referencing==0.30.0
regex==2023.10.3
requests==2.27.1
rich==13.6.0
rlp==3.0.0
rpds-py==0.9.2
s3transfer==0.5.2
six==1.16.0
SQLAlchemy==1.4.47
SQLAlchemy-Utils==0.41.1
toolz==0.12.0
typing_extensions==4.7.1
urllib3==1.26.16
varint==1.0.2
wcwidth==0.2.6
web3==6.10.0
websockets==12.0
yarl==1.9.2
zipp==3.17.0

What was wrong?

Calling eth.contract() frequently and constantly causes memory usage increase a lot.

  def contract(self, contract, abi):
      start = time.time()
      try:
          self.api_count_tracking[CONTRACT_CALL] += 1
          c = self.w3_client.eth.contract(safe_to_check_sum_address(contract), abi=abi)
          end = time.time()
          self.time_tracking[CONTRACT_CALL] += end - start
          return c
      except Exception as e:
          logging.error(f'contract not found: contract_address is {contract} error is {e}. abi: {abi}')
          end = time.time()
          self.time_tracking[CONTRACT_CALL] += end - start

We call this function quite frequently, over 100 times/sec using multithreading and multiprocesses.
I used memory profiler to analyze it and it shows that ContractFactory = ContractFactoryClass.factory(self.w3, **kwargs) was allocated a lot of memory. Another candidate could be venv/lib/python3.9/site-packages/web3/_utils/datatypes.py:61

I found one issue mentioning this problem here: #2296

How can it be fixed?

Not sure how to fix that.

---

Note: We prefer to use issues to track our work. If you think you've encountered a bug in web3py or
have a feature request, you're in the right place. If you have implementation or usage questions,
please refer to our documentation and/or join the conversation
on discord.

[pacrob]

Being tracked in #2296