I'm concerned that I'm seeing an issue similar to this old one 
<https://github.com/grpc/grpc/issues/15889>, so I want to verify that 
`TCP_USER_TIMEOUT` is set to something reasonable.

To configure that value, it sounds like gRPC keepalive must be enabled, but 
there are concerns this is "dangerous" 
<https://github.com/grpc/grpc/issues/25713>. Additionally, there appears to 
be at least one bug <https://github.com/grpc/grpc/issues/36411> in the 
Python keepalive configuration. Is it possible to side-step these issues by 
configuring the `TCP_USER_TIMEOUT` separately from the grpc keepalive?

Additionally, is it possible to verify how the timeout is configured? From 
the source code 
<https://github.com/grpc/grpc/blob/master/src/core/lib/iomgr/socket_utils_common_posix.cc#L367>,
 
looks like there should be log messages explicitly indicating whether 
`TCP_USER_TIMEOUT` is supported and how it's configured. However, I don't 
see any log entries containing that string, even though I do see logging 
from `tcp_posix.cc`.

I have these environment variables set:

```
GRPC_VERBOSITY="INFO"
GRPC_TRACE="tcp,http_keepalive,http2_ping,resource_quota,queue_pluck,timer,health_check_client,executor,connectivity_state"
```

...but I don't see any log messages with `TCP_USER_TIMEOUT`.

-- 
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 [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/grpc-io/b98fb1ac-f35d-486a-adfe-3c6309ed82fdn%40googlegroups.com.

Reply via email to