[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 271069] syslogd service inside client jail requires restart before server jail receives logs"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 25 Apr 2023 19:44:13 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=271069 Bug ID: 271069 Summary: syslogd service inside client jail requires restart before server jail receives logs Product: Base System Version: 13.2-RELEASE Hardware: Any OS: Any Status: New Severity: Affects Only Me Priority: --- Component: bin Assignee: bugs@FreeBSD.org Reporter: me@svmhdvn.name I'm not sure if this is an actual bug or an issue with my system configuration: PROBLEM: In a centralized logging configuration with one client jail (sending logs) and one server jail (receiving logs and writing them to local files), syslogd inside the *client* jail requires a restart before I can observe on the *server* jail that the logs are successfully received. This is observed on a fresh start of the *server* jail, followed by a fresh start of the *client* jail. Is there a race condition or order of operations problem somewhere? Let me know if I need to supply more info about the configuration. DETAILS: I have a simple configuration of two standard thick jails (named 'ssh' and 'logs') with this configuration: ===== /etc/jail.conf ===== mount.devfs; allow.raw_sockets; exec.clean; exec.timeout = 30; stop.timeout = 30; path = "/usr/jail/guests/${name}"; host.hostname = "${name}.my.domain"; exec.start = "/bin/sh /etc/rc"; exec.stop = "/bin/sh /etc/rc.shutdown"; logs { ip6.addr = "re0|fdac::aaaa:201/64"; } ssh { ip6.addr = "re0|fdac::aaaa:202/64"; depend = "logs"; } ========================== Each jail has the same content in their host files and no DNS servers (i.e. name resolution is available through hosts files only): ===== /etc/hosts ===== [...] fdac::aaaa:201 logs.my.domain logs fdac::aaaa:202 ssh.my.domain ssh [...] ====================== In the client jail ssh.my.domain: ===== /etc/syslog.conf ==== *.* @logs.my.domain =========================== ===== /etc/rc.conf ==== syslogd_flags="-8 -O syslog -s -v -v" =========================== In the server jail logs.my.domain: ===== /etc/syslog.conf ==== +ssh.my.domain *.* /var/log/ssh.my.domain.log =========================== ===== /etc/rc.conf ==== syslogd_flags="-8 -O syslog -a '*.my.domain' -v -v" =========================== Steps to repro (as root): 1. in the jailhost: # service jail onestart logs # service jail onestart ssh 2. inside logs.my.domain (for observing received logs): # tail -f /var/log/ssh.my.domain.log [...] follow the log 3. inside ssh.my.domain: # logger "hello from ssh" <--- *NOT* observed on the log server # service syslogd restart # logger "hello from ssh" <--- SUCCESS visible in log server -- You are receiving this mail because: You are the assignee for the bug.