Attention is currently required from: pespin, fixeria, daniel.
2 comments:
Patchset:
But if that's unrelated to this patch and it shows up as a result of rebase, better submit a precedi […]
I have added a patch eliminating timestamp logging in the test:
https://gerrit.osmocom.org/c/libosmo-netif/+/34224
I don't know what the cause is, but it appears recent changes to libosmocore or libosmo-netif have changed the timestamp behavior which used to be predictable.
Patchset:
After adapting `stream_test.err` to some of the new logs (locally), a connectivity issue (it's not just the logging, a lag that I didn't have before comes along with the following output) has appeared
```
unable to connect socket: 127.0.0.11:1111: Cannot assign requested address
no suitable remote addr found for: 127.0.0.11:1111
CLICONN(,){WAIT_RECONNECT} retrying reconnect in 5 seconds...
Cannot open stream client
```
It's not any patches in libosmocore causing this new behavior (I tried an older libosmocore patch, that didn't change anything), so I took a look at libosmo-netif and saw a couple of recent changes, probably something there.
A workaround for this was changing the port for the stream client tests from 1111 to 1112 (similarly to how I had to change the local port from 8967 to something else in order for osmo time not becoming unpredictable...).
But I do wonder if this shouldn't be addressed (somehow after destroying the client, and reopening the server link, the same port remains blocked for a noticeable amount of time, apart from the error log).
To view, visit change 33198. To unsubscribe, or for help writing mail filters, visit settings.