# No Logging Options for Serve Deployment and/or Ray generated Pool Actors

**URL:** <https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164>\
**Category:** Ray Serve\
**Created:** [February 22, 2022, 9:49pm UTC](https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164 "2022-02-22T21:49:21Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Yatagarasu](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/yatagarasu/32/2192_2.png) [@Yatagarasu](https://discuss.ray.io/u/Yatagarasu)\
**Post date:** [February 22, 2022, 9:49pm UTC](https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164/1 "2022-02-22T21:49:22Z")

</div>

Salutations,

I am looking for an option(s) for controlling the logging level for a serve deployment and pool actors (generated under the Ray Pool API) with Python and have not been able to do so, in either the Ray source code or existing documentation/issues. Specifying the logging level at initialization should pass the level to any actors using the encapsulated resources, though this doesn’t actually appear to be the case. A minimum example code for reproduction of the serve deployment is included below (though similar behavior occurs with pool actors as described/exhibited). Any insights would be appreciated; thank you in advanced.

**Versions / Dependencies**

> Windows 10  
> Ubuntu 18.04  
> Python 3.8  
> Ray 1.10.0

**Reproduction script**

```auto
import logging
import ray
from ray import serve
import time
  
ray.init(configure_logging=True, logging_level=logging.ERROR)
  
@serve.deployment(route_prefix="/ModelServer")
class ModelServer:
    def __init__ (self): print('Initializing')
    def __call__ (self): print('Called')

serve.start()
ModelServer.deploy()
model = ModelServer.get_handle()
ray.get(model.remote())
serve.shutdown()

```

**What happened**

```auto
 pid=47948) 2022-02-22 15:30:47,910 INFO checkpoint_path.py:16 -- Using RayInternalKVStore for controller checkpoint and recovery.
 pid=47948) 2022-02-22 15:30:48,017 INFO http_state.py:98 -- Starting HTTP proxy with name 'SERVE_CONTROLLER_ACTOR:mQCdAu:SERVE_PROXY_ACTOR-node:127.0.0.1-0' on node 'node:127.0.0.1-0' listening on '127.0.0.1:8000'
2022-02-22 15:30:48,655 INFO api.py:475 -- Started Serve instance in namespace '0d3370fc-79eb-44c2-9487-469ea0e95e50'.
2022-02-22 15:30:48,670 INFO api.py:249 -- Updating deployment 'ModelServer'. component=serve deployment=ModelServer
 pid=2228) INFO: Started server process [2228]
 pid=47948) 2022-02-22 15:30:48,769 INFO deployment_state.py:920 -- Adding 1 replicas to deployment 'ModelServer'. component=serve deployment=ModelServer
2022-02-22 15:30:49,437 INFO api.py:261 -- Deployment 'ModelServer' is ready at `http://127.0.0.1:8000/ModelServer`. component=serve deployment=ModelServer
 pid=40024) Initializing
 pid=40024) Called
 pid=47948) 2022-02-22 15:30:49,542 INFO deployment_state.py:940 -- Removing 1 replicas from deployment 'ModelServer'. component=serve deployment=ModelServer

```

**What you expected to happen**

```auto
 pid=40024) Initializing
 pid=40024) Called

```

---

<div class="post-metadata">

**Author:** ![architkulkarni](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/architkulkarni/32/10_2.png) [@architkulkarni](https://discuss.ray.io/u/architkulkarni)\
**Post date:** [February 22, 2022, 11:16pm UTC](https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164/2 "2022-02-22T23:16:41Z")

</div>

Hi @Yatagarasu, thanks for reporting this! I agree, I would expect that setting the logging level as you did would hide the INFO logs. @sangcho can I confirm with you that this is how the `configure_logging` and `logging_level` APIs are supposed to work?

If so, this looks like a bug in Ray Serve (and possibly also in Ray Pool), and we can make a Github issue for this to track it.

---

<div class="post-metadata">

**Author:** ![Yatagarasu](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/yatagarasu/32/2192_2.png) [@Yatagarasu](https://discuss.ray.io/u/Yatagarasu)\
**Post date:** [February 23, 2022, 7:59pm UTC](https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164/3 "2022-02-23T19:59:06Z")

</div>

Thank you for your reply; now opened as an issue:

> <https://github.com/ray-project/ray/issues/22605>
>
> \### Search before asking
> 
> \- \[X\] I searched the \[issues\](https://github.com/ray-p…roject/ray/issues) and found no similar issues.
> 
> 
> \### Ray Component
> 
> Ray Serve
> 
> \### What happened + What you expected to happen
> 
> Re-posted from: https://discuss.ray.io/t/no-logging-options-for-serve-deployment-and-or-ray-generated-pool-actors/5164 
> 
> Option(s) for controlling the logging level for a serve deployment and pool actors (generated under the Ray Pool API) with Python are not functional. Specifying the logging level at initialization should pass the specified level to any actors using the encapsulated resources, though this doesn’t actually appear to be the case. Reproduction script included for serve deployment, though similar behavior also occurs with pool actors. Thank you in advanced.
> 
> \*\*What happened\*\*
> \`\`\`
> pid=47948) 2022-02-22 15:30:47,910 INFO checkpoint\_path.py:16 -- Using RayInternalKVStore for controller checkpoint and recovery.
> pid=47948) 2022-02-22 15:30:48,017 INFO http\_state.py:98 -- Starting HTTP proxy with name 'SERVE\_CONTROLLER\_ACTOR:mQCdAu:SERVE\_PROXY\_ACTOR-node:127.0.0.1-0' on node 'node:127.0.0.1-0' listening on '127.0.0.1:8000'
> 2022-02-22 15:30:48,655 INFO api.py:475 -- Started Serve instance in namespace '0d3370fc-79eb-44c2-9487-469ea0e95e50'.
> 2022-02-22 15:30:48,670 INFO api.py:249 -- Updating deployment 'ModelServer'. component=serve deployment=ModelServer
> pid=2228) INFO: Started server process \[2228\]
> pid=47948) 2022-02-22 15:30:48,769 INFO deployment\_state.py:920 -- Adding 1 replicas to deployment 'ModelServer'. component=serve deployment=ModelServer
> 2022-02-22 15:30:49,437 INFO api.py:261 -- Deployment 'ModelServer' is ready at \`http://127.0.0.1:8000/ModelServer\`. component=serve deployment=ModelServer
> pid=40024) Initializing
> pid=40024) Called
> pid=47948) 2022-02-22 15:30:49,542 INFO deployment\_state.py:940 -- Removing 1 replicas from deployment 'ModelServer'. component=serve deployment=ModelServer
> \`\`\`
> 
> \*\*What you expected to happen\*\*
> \`\`\`
> pid=40024) Initializing
> pid=40024) Called
> \`\`\`
> 
> 
> \### Versions / Dependencies
> 
> \`\`\`
> Windows 10 and/or Ubuntu 18.04
> Python 3.8
> Ray 1.10.0
> \`\`\`
> 
> \### Reproduction script
> 
> \`\`\`
> import logging
> import ray
> from ray import serve
>   
> ray.init(configure\_logging=True, logging\_level=logging.ERROR)
>   
> @serve.deployment(route\_prefix="/ModelServer")
> class ModelServer:
> def \_\_init\_\_(self): print('Initializing')
> def \_\_call\_\_(self): print('Called')
> 
> serve.start()
> ModelServer.deploy()
> model = ModelServer.get\_handle()
> ray.get(model.remote())
> serve.shutdown()
> \`\`\`
> 
> \### Anything else
> 
> \_No response\_
> 
> \### Are you willing to submit a PR?
> 
> \- \[\] Yes I am willing to submit a PR!