Re: [grpc-io] gRPC python server memory consumption

['Lidi Zheng' via grpc.io]

Can you create a reproducible example and post it to
https://github.com/grpc/grpc/issues?

On Mon, Apr 20, 2020 at 6:47 PM JS  wrote:

> Hi,
>
> I am currently using gRPC python server and seeing an issue.
>
> This is with gRPC
> *Server*: Python
> *Client*: go-lang
>
> There are Unary RPCs defined in the python gRPC server which basically
> execute an ansible task (ansible spawns a subprocess forking off as a child
> from the main top level process) and returns the result to the end user.
> This task can be time consuming.
>
> I have defined a callback to handle any client side cancellations on
> timeout.
>
> Right now, I see there is a lot of memory that accumulates over time after
> running these RPCs which doesn't get freed once rpc is done.
>
> Do we need to handle any server side clean up after executing the RPC?
>
> The leak is significant that its impacting the functionality. often see *"err:
> rpc error: code = DeadlineExceeded desc = Deadline Exceeded" *when client
> sends a heartbeat request to the server.
>
> Using python 3.6 and gRPC version 1.24
>
> The memory goes up to ~6gb over time before OOM killer kicks in.
>
> Thanks
> J/-
>
> --
> You received this message because you are subscribed to the Google Groups "
> grpc.io" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to grpc-io+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/grpc-io/7b3f041d-e9be-4793-a7af-8614b5431a74%40googlegroups.com
> 
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"grpc.io" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to grpc-io+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/grpc-io/CAMC1%3Dje%3D38jGK9jcULmX1g2gQ-k7PGf0vZU%3DF8ndD9-Fj_28yg%40mail.gmail.com.

[grpc-io] gRPC python server memory consumption

[JS]

Hi,

I am currently using gRPC python server and seeing an issue. 

This is with gRPC 
*Server*: Python
*Client*: go-lang 

There are Unary RPCs defined in the python gRPC server which basically 
execute an ansible task (ansible spawns a subprocess forking off as a child 
from the main top level process) and returns the result to the end user. 
This task can be time consuming.

I have defined a callback to handle any client side cancellations on 
timeout. 

Right now, I see there is a lot of memory that accumulates over time after 
running these RPCs which doesn't get freed once rpc is done. 

Do we need to handle any server side clean up after executing the RPC?

The leak is significant that its impacting the functionality. often see *"err: 
rpc error: code = DeadlineExceeded desc = Deadline Exceeded" *when client 
sends a heartbeat request to the server.

Using python 3.6 and gRPC version 1.24

The memory goes up to ~6gb over time before OOM killer kicks in.

Thanks
J/-

-- 
You received this message because you are subscribed to the Google Groups 
"grpc.io" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to grpc-io+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/grpc-io/7b3f041d-e9be-4793-a7af-8614b5431a74%40googlegroups.com.