Open nyxdata opened 2 years ago
Windows deployment and updates are 5-15x slower on Windows than on Linux and macOS unfortunately. Zappa was developed on a mac and works best on UNIX-based.
I tried debugging it when I was on Windows, but it was a problem with a recursive function which goes through the folders of the project and it was not easy to pin-point the problem
Before I changed to mac, I was using WSL2 to "bypass" that issue, which I recommend you using too
I'm on ubuntu, zappa is also running extremely slow on my laptop , and it used to be very fast, it uses 100% CPU
$ zappa -v
0.57.0
$:~/talkToGiantSummary$ top
top - 22:12:05 up 29 min, 1 user, load average: 1.48, 1.48, 0.88
Tasks: 330 total, 2 running, 328 sleeping, 0 stopped, 0 zombie
%Cpu(s): 8.2 us, 0.4 sy, 0.0 ni, 90.9 id, 0.4 wa, 0.0 hi, 0.0 si, 0.0 st
MiB Mem : 15641.4 total, 192.3 free, 3171.5 used, 12277.6 buff/cache
MiB Swap: 2048.0 total, 1972.0 free, 76.0 used. 10707.0 avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
3417 xuan 20 0 191676 101900 12360 R 95.3 0.6 7:53.11 zappa
Hi @Techget You need to provide your zappa_settings.json file and requirements.txt file in order for us to pinpoint the issue.
Thanks for the reminder, following is the corresponding info
zappa_seetings.json:
{
"dev": {
"app_function": "api.app",
"aws_region": "ap-southeast-2",
"profile_name": "default",
"project_name": "project name",
"runtime": "python3.8",
"s3_bucket": "s3_bucket",
"lambda_description": "lambda to generate ttg summary",
"apigateway_description": "Serverless API endpoints using Zappa for lambda to generate ttg summary",
"cors": true,
"slim_handler": true,
"timeout_seconds": 300,
"environment_variables": {
}
}
}
requirements.txt
aiohttp==3.8.4
aiosignal==1.3.1
argcomplete==3.1.1
async-timeout==4.0.2
boto3==1.26.155
botocore==1.29.155
Flask==2.0.2
flask-core==2.9.0
Flask-Cors==3.0.10
Flask-SQLAlchemy==2.5.1
itsdangerous==2.0.1
Jinja2==3.0.3
sentence-transformers==2.2.2
networkx==3.1
scikit-learn==1.2.2
langchain==0.0.202
langchainplus-sdk==0.0.10
numexpr==2.8.4
numpy==1.25.0
openai==0.27.8
openapi-schema-pydantic==1.2.4
packaging==23.1
Pillow==9.5.0
placebo==0.9.0
psycopg2==2.9.2
psycopg2-binary==2.9.2
pycparser==2.21
PyMySQL==1.0.2
pyparsing==3.1.0
PyYAML==6.0
regex==2023.6.3
requests==2.31.0
s3transfer==0.6.1
scipy==1.11.0
six==1.16.0
SQLAlchemy==1.4.28
sympy==1.12
threadpoolctl==3.1.0
toml==0.10.2
tqdm==4.65.0
triton==2.0.0
troposphere==4.3.2
typing-inspect==0.9.0
typing_extensions==4.6.3
urllib3==1.26.16
Werkzeug==2.0.2
yarl==1.9.2
zappa==0.57.0
So, you have only one region deployment or multiple region deployments? If it's only one then I would look into the internet connection and the place from where you're trying to deploy.
Deploying to single region and the network speed is okay with ~5 MB/s.
It often stucks at a stage for quite while e.g. Packaging project as gzipped tarball.
Here is what I would do:-
Let me know what worked for you. I will try to replicate your setup.
You can also join our Slack channel using this link
Hi there! Unfortunately, this Issue has not seen any activity for at least 90 days. If the Issue is still relevant to the latest version of Zappa, please comment within the next 10 days if you wish to keep it open. Otherwise, it will be automatically closed.
still have the same issue.
I've built a small serverless application about ~10mo.
When I use the command zappa update, it takes at least 10 minutes before the beginning of the update.
I use windows and the following settings:
{ "dev": { "app_function": "app.app", "aws_region": "eu-west-1", "profile_name": "user", "project_name": "nyx", "runtime": "python3.8", "s3_bucket": "zappa-nyw", "memory_size": 2048 } }
I've tried to add exclusion exceptions to get it faster but nothing has changed.
Thank you for your help.