Dennis Felsing
Dennis Felsing
Anyone can reproduce this?
I'm actually seeing this error on all servers we are running, not just GER1.
Also seeing lots of these, I guess they actually cause the data loss: ``` [2021-12-05 20:25:36][sql]: WARNING: previous save score result didn't complete, overwriting it now ```
The guess of what happened on the server is that packets with an mss > 1400 were blocked. I guess this lead to database connections kind of staying alive, but...
That would be great, thanks!
We still have quite some failures like this: ``` [2022-05-25 13:35:28][mysql]: can't deinitialize, connections remaining: 2 ``` Anything we can do to better analyze what's happening? I'm just killing them...
@Zwelf would it help to get a gdb session when this fails?
On some servers every few days, on most others not. Edit: Actually now that I check it's happening on most servers, and a lot.
``` [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". __pthread_clockjoin_ex (threadid=140254934619904, thread_return=thread_return@entry=0x0, clockid=clockid@entry=0, abstime=abstime@entry=0x0, block=block@entry=true) at pthread_join_common.c:145 145 pthread_join_common.c: No such file or directory. (gdb) thread apply all...
@heinrich5991 On all I see the master server connection still running with curl_multi_poll, could this be related? Edit: I see this change was added shortly before: https://github.com/ddnet/ddnet/pull/5202