Marc Huber

Hi, the "Connection refused" message at this place is the error code returned by sendmsg(), likely indicating that the other side of the connection is no longer running. Were there...

Hi, the listener component accepts connections and forwards them (via UNIX domain sockets) to the worker processes. Apparently, one of the worker processes was no longer able to handle the...

Hi, thanks for reporting. This apparently happens somewhere in the error path after spawnd_accepted.c:422. I really can't read much more from the syslog output, compiling with --debug or --ggdb and...

Hi, I've just commited 443023f75536a217db5f5216c828775a0b5276b8, no idea whether that fixes the issue, this is just based on guessing. Thanks, Marc

Hi, that memory leak is clearly unexpected. Could you please share some details about your configuration (mavis modules used, e.g.)? Thanks, Marc

Hi, thanks, that should make it easier to find what's causing this. Might take some time, nevertheless, Thanks, Marc

Hi, do logs actually get written to /var/log/tac_plus/...? That location being non-writable would be a fairly good excuse for memory filling up. Just guessing. Thanks, Marc

Hi, thanks, so I'll look further into finding that memory leak :-( The "single-connection" warning should go away after defining "single-connection = yes" at host or realm level. Thanks, Marc

Hi, please forcefully "kill -9 3176288 3176289" (or whatever the process ids are now), 1474064 and 3176288/3176289 don't look related to me. After that, please restart tac_plus-ng (and make certain...

Hi, I simply can't test that, due to lack of suitable hardware. May or may not be an issue with IOS-XE, or with tac_plus-ng. In general, if restarting tac_plus-ng causes...