11 Sep
2002
11 Sep
'02
6:34 p.m.
Ted_Rule@flextech.co.uk writes:
I previously enquired whether it would be possible to allow for a keep-alive option on Unix DGRAM sockets to help avoid the situation on some platforms when daemons such as sendmail have the /dev/log socket 'whipped from under them' during syslog-ng/SIGHUP.
Why not fix the libc code on those platforms so it does basic sanity checking on the socket, and re-opens the socket when it discovers that its cached filehandle for /dev/log is invalid? I encountered this on NetBSD and it was an easy library fix (it's patched in all current releases (1.5.3 and 1.6 candidates) - what OS are you using?