fyoorer / ShadowClone

Unleash the power of cloud
Apache License 2.0
718 stars 95 forks source link

Increase memory #69

Open marcelo321 opened 1 month ago

marcelo321 commented 1 month ago

I want to increase memory from 1024 to 2048.

2024-08-29 20:01:54,853 [INFO] invokers.py:108 -- ExecutorID 5c412a-0 | JobID M000 - Selected Runtime: sc-runtime - 1024MB
2024-08-29 20:02:01,110 [INFO] invokers.py:172 -- ExecutorID 5c412a-0 | JobID M000 - Starting function invocation: execute_command_chunks() - Total: 399 activations
2024-08-29 20:02:03,270 [INFO] invokers.py:208 -- ExecutorID 5c412a-0 | JobID M000 - View execution logs at /tmp/lithops-root/logs/5c412a-0-M000.log
2024-08-29 20:02:03,362 [INFO] wait.py:97 -- ExecutorID 5c412a-0 - Getting results from 399 function activations

I thought it was configurable with lithops runtime deploy sc-runtime --memory 1024 --timeout 900 but it isn't taking effect.

I try:

lithops runtime build sc-runtime --memory 2048 lithops runtime delete sc-runtime --memory 1024

but then

# lithops runtime list
2024-08-29 20:08:56,659 [INFO] config.py:134 -- Lithops v2.8.0
2024-08-29 20:08:56,907 [INFO] aws_lambda.py:94 -- AWS Lambda client created - Region: us-east-1

Runtime Name     Memory Size     Lithops Version
----------       -------------   -----------------
sc-runtime       2048 MB         2.8.0
sc-runtime       1024 MB         2.8.0

Total runtimes: 2

still 1024 and crashes since I have no image of 1024 (I manually deleted it):

2024-08-29 20:35:44,762 [INFO] Uploading NOSPLIT file to storage as-is
2024-08-29 20:35:44,763 [INFO] Uploading NOSPLIT file to storage as-is
2024-08-29 20:35:44,763 [INFO] Large input file provided. Using file chunks
2024-08-29 20:35:44,763 [INFO] Splitting input file into chunks of 635 lines
2024-08-29 20:35:44,980 [INFO] Uploading chunks to storage
2024-08-29 20:35:45,081 [INFO] config.py:134 -- Lithops v2.8.0
2024-08-29 20:35:45,088 [INFO] aws_s3.py:60 -- S3 client created - Region: us-east-1
2024-08-29 20:35:45,231 [INFO] aws_lambda.py:94 -- AWS Lambda client created - Region: us-east-1
2024-08-29 20:35:45,234 [INFO] invokers.py:108 -- ExecutorID a01f8b-0 | JobID M000 - Selected Runtime: sc-runtime - 1024MB
2024-08-29 20:35:45,351 [INFO] invokers.py:116 -- Runtime sc-runtime with 1024MB is not yet deployed
2024-08-29 20:35:45,351 [INFO] aws_lambda.py:455 -- Deploying runtime: sc-runtime - Memory: 1024 Timeout: 900

EDIT: ooh I needed to edit nano ~/.lithops/config as well