Instance performance degradation

Question

I performed a stop and start on the application and as a result, the instance normality became constantly degraded and the application stopped working. After checking the logs, I am getting the following error.

I would like to know how to remedy this problem.

#Log error
2024-06-22T02:14:12.805Z ERROR - Container thinxpand_0_ecc28606 for site thinxpand has exited, failing site start

2024-06-22T02:14:12.807Z INFO - Initiating warmup request to container thinxpand_0_ecc28606_middleware for site thinxpand

2024-06-22T02:14:13.060Z INFO - Container thinxpand_0_ecc28606_middleware for site thinxpand initialized successfully and is ready to serve requests.

2024-06-22T02:14:13.091Z ERROR - Container thinxpand_0_ecc28606 didn't respond to HTTP pings on port: 8000, failing site start. See container logs for debugging.

2024-06-22T02:14:13.820Z INFO - Stopping site thinxpand because it failed during startup.

Answer 1

@涌井大輔 Thanks for your patience on this.

I have reviewed the backend logs of the above shared app service "Thin**pand" and we can see that site is failing to start with the below error message.

ValueErro* numpy.dty* size changed, may indicate binary incompati* Expected 96 from C header, got 88 from PyObject

Based on the error message and as mentioned in the Similar SO thread here. Requested Op to downgrade the NumPy version to 1.26.4.

Op has tried and confirmed that they are able to start the app service post downgrading.

Feel free to reach back to me if you have any further questions on this.