Estou tentando estabelecer uma conexão entre os serviços firewalld e docker. Estendi o arquivo de unidade firewalld.service padrão adicionando o seguinte:
[Service]
ExecStartPost=-/bin/bash -c '/usr/bin/test -f /etc/default/docker && /bin/systemctl stop docker && /bin/systemctl start docker'
para /etc/systemd/system/firewalld.service.d/docker.conf
.
Mas isso não funcionou, o serviço firewalld expirou no início:
root@ubuntu-xenial:~# systemctl status firewalld
● firewalld.service - firewalld - dynamic firewall daemon
Loaded: loaded (/lib/systemd/system/firewalld.service; enabled; vendor preset: enabled)
Drop-In: /etc/systemd/system/firewalld.service.d
└─docker.conf
Active: failed (Result: timeout) since Sat 2017-04-08 14:39:45 UTC; 1min 35s ago
Process: 26050 ExecStartPost=/bin/bash -c /usr/bin/test -f /etc/default/docker && /bin/systemctl stop docker && /bin/systemctl start docker (code=killed, signal=TERM)
Process: 26000 ExecStart=/usr/sbin/firewalld --nofork --nopid (code=exited, status=0/SUCCESS)
Main PID: 26000 (code=exited, status=0/SUCCESS)
Apr 08 14:38:10 ubuntu-xenial systemd[1]: Starting firewalld - dynamic firewall daemon...
Apr 08 14:39:41 ubuntu-xenial systemd[1]: firewalld.service: Start-post operation timed out. Stopping.
Apr 08 14:39:45 ubuntu-xenial systemd[1]: Failed to start firewalld - dynamic firewall daemon.
Apr 08 14:39:45 ubuntu-xenial systemd[1]: firewalld.service: Unit entered failed state.
Apr 08 14:39:45 ubuntu-xenial systemd[1]: firewalld.service: Failed with result 'timeout'.
Estou ciente disso BindTo
, Requires
mas gostaria de reiniciar o serviço docker a cada mudança de estado do firewalld para que essas diretivas não resolvam meu problema.
Upd.: strace
mostra que systemctl start
está esperando em um soquete unix:
root@ubuntu-xenial:~# lsof -p 27898
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
systemctl 27898 root cwd DIR 8,1 4096 2 /
systemctl 27898 root rtd DIR 8,1 4096 2 /
systemctl 27898 root txt REG 8,1 659848 124 /bin/systemctl
systemctl 27898 root mem REG 8,1 80496 2057 /lib/x86_64-linux-gnu/libgpg-error.so.0.17.0
systemctl 27898 root mem REG 8,1 14608 2060 /lib/x86_64-linux-gnu/libdl-2.23.so
systemctl 27898 root mem REG 8,1 456632 2144 /lib/x86_64-linux-gnu/libpcre.so.3.13.2
systemctl 27898 root mem REG 8,1 1864888 2066 /lib/x86_64-linux-gnu/libc-2.23.so
systemctl 27898 root mem REG 8,1 138696 2082 /lib/x86_64-linux-gnu/libpthread-2.23.so
systemctl 27898 root mem REG 8,1 919168 2037 /lib/x86_64-linux-gnu/libgcrypt.so.20.0.5
systemctl 27898 root mem REG 8,1 137400 2067 /lib/x86_64-linux-gnu/liblzma.so.5.0.0
systemctl 27898 root mem REG 8,1 31712 2077 /lib/x86_64-linux-gnu/librt-2.23.so
systemctl 27898 root mem REG 8,1 130224 2154 /lib/x86_64-linux-gnu/libselinux.so.1
systemctl 27898 root mem REG 8,1 162632 2069 /lib/x86_64-linux-gnu/ld-2.23.so
systemctl 27898 root 0r CHR 1,3 0t0 6 /dev/null
systemctl 27898 root 1w CHR 1,3 0t0 6 /dev/null
systemctl 27898 root 2w CHR 1,3 0t0 6 /dev/null
systemctl 27898 root 3u unix 0xffff880003155400 0t0 177152 type=STREAM
root@ubuntu-xenial:~# strace -p 27898
strace: Process 27898 attached
ppoll([{fd=3, events=POLLIN}], 1, NULL, NULL, 8 <unfinished ...>
+++ killed by SIGKILL +++
root@ubuntu-xenial:~#
Atualizar 2: de acordo com gdb
, systemctl
está bloqueado em uma chamada DBus:
root@ubuntu-xenial:~# gdb -ex "set pagination 0" -ex "thread apply all bt" --batch -p 18145
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
0x00007f99b3edebcd in __GI_ppoll (fds=0x7ffc098bb800, nfds=1, timeout=<optimized out>, sigmask=0x0) at ../sysdeps/unix/sysv/linux/ppoll.c:50
50 ../sysdeps/unix/sysv/linux/ppoll.c: No such file or directory.
Thread 1 (Thread 0x7f99b4f0d880 (LWP 18145)):
#0 0x00007f99b3edebcd in __GI_ppoll (fds=0x7ffc098bb800, nfds=1, timeout=<optimized out>, sigmask=0x0) at ../sysdeps/unix/sysv/linux/ppoll.c:50
#1 0x000056094741272e in ppoll () at /usr/include/x86_64-linux-gnu/bits/poll2.h:71
#2 bus_poll.lto_priv.198 (bus=0x560948d86040, need_more=<optimized out>, timeout_usec=<optimized out>) at ../src/libsystemd/sd-bus/sd-bus.c:2874
#3 0x00005609473b435e in bus_process_wait (bus=0x560948d86040) at ../src/shared/bus-util.c:1993
#4 bus_wait_for_jobs (extra_args=0x0, quiet=false, d=0x560948d86910) at ../src/shared/bus-util.c:2136
#5 start_unit.lto_priv.362 (argc=<optimized out>, argv=<optimized out>, userdata=<optimized out>) at ../src/systemctl/systemctl.c:2787
#6 0x00005609473b0716 in dispatch_verb (verbs=0x56094743f760 <verbs>, userdata=0x0, argv=0x7ffc098bbb38, argc=<optimized out>) at ../src/basic/verbs.c:92
#7 systemctl_main () at ../src/systemctl/systemctl.c:7447
#8 main (argc=<optimized out>, argv=0x7ffc098bbb38) at ../src/systemctl/systemctl.c:7713
root@ubuntu-xenial:~#
Consegui resolver esse problema adicionando
--no-block
uma opção aosystemctl
comando assim: