当蛤lam运行时，服务器在重新启动时挂起

简而言之

重新启动通常需要不到1分钟的时间，但是如果clamav-daemon正在运行，则重新启动将完全需要30分钟。

概述

我有一个Ubuntu 15.10“云”邮件服务器：postfix，dovecot，mysql，可通过SSH访问。这已经工作了好几个星期，并在不到1分钟的时间内重启。一切都很好，直到我通过Ubuntu的apt-get安装软件包安装了Amavis－现在每次重新启动仅花费30分钟。

运行时，一切正常。我检查了日志，没有明显的问题。Clamav正在工作，邮件收发，CPU使用率在0％到0.5％之间波动。

我的理论

我已将问题缩小到clamav-daemon消除的过程。唯一的新安装是Amavis，它添加​​了新的守护程序。如果我都停止了它们（amavis，clamav-daemon和clamav-freshclam），则重新启动是即时的。如果我只是停止clamav-daemon，重新启动是即时的。但是，如果我继续运行并停止所有其他运行，则重新启动又需要30分钟。

这是完全可重复的，等待总是恰好30分钟。这让我认为Clamav正在等待用户输入，超时时间为30分钟。

我可以看到Clamav的进程正在使用“ --foreground=true”运行。这使我认为Amavis故意这样做是为了捕获输出，但是不希望在重新启动过程中出现提示。

从htop：

我的日志

注意：/etc/init.d/sendsigs已report_unkillable启用以捕获不可杀死的进程，并且/etc/default/apport已启用。/var/crash最近没有登录。

（大约在14:51左右重新启动，并且直到15:23-30分钟后服务器才准备好SSH，之间没有错误。）

/var/log/clamav/clamav.log

...
14:52:45 2016 -> --- Stopped at Fri 14:52:45 2016
15:23:04 2016 -> +++ Started at Fri 15:23:04 2016
15:23:04 2016 -> Received 1 file descriptor(s) from systemd.
15:23:04 2016 -> clamd daemon 0.98.7 (OS: linux-gnu, ARCH: x86_64, CPU: x86_64)
15:23:04 2016 -> Running as user clamav (UID 119, GID 127)
...

/var/log/mail.log

...
14:51:15 mail amavis[1578]: Using primary internal av scanner code for ClamAV-clamd
14:51:15 mail amavis[1578]: Found secondary av scanner ClamAV-clamscan at /usr/bin/clamscan
14:51:15 mail amavis[1578]: Deleting db files __db.001,nanny.db,__db.003,snmp.db,__db.002 in /var/lib/amavis/db
14:51:15 mail amavis[1578]: Creating db in /var/lib/amavis/db/; BerkeleyDB 0.55, libdb 5.3
15:23:05 mail postgrey[820]: Process Backgrounded
15:23:05 mail postgrey[820]: 2016/02/12-15:23:05 postgrey (type Net::Server::Multiplex) starting! pid(820)
15:23:05 mail postgrey[820]: Resolved [localhost]:10023 to [::1]:10023, IPv6
15:23:05 mail postgrey[820]: Resolved [localhost]:10023 to [127.0.0.1]:10023, IPv4
...

/ var / log / syslog

...
14:52:18 mail systemd[1]: Stopping ClamAV virus database updater...
14:52:18 mail freshclam[760]: Update process terminated
14:52:18 mail systemd[1]: Stopped ClamAV virus database updater.
14:52:42 mail systemd[1]: Stopped Setup Virtual Console.
14:52:42 mail systemd[1]: Stopping LVM2 PV scan on device 202:2...
14:52:42 mail systemd[1]: Deactivating swap /dev/disk/by-uuid/758b48d5-df40-4ca8-af0d-a482044d21dd...
14:52:42 mail systemd[1]: Stopping Authenticate and Authorize Users to Run Privileged Tasks...
14:52:42 mail systemd[1]: Stopping User Manager for UID 1000...
14:52:42 mail systemd[1907]: Reached target Shutdown.
14:52:42 mail swapoff[2140]: swapoff: /dev/mapper/ubuntu-swap: swapoff failed: Cannot allocate memory
14:52:42 mail systemd[1907]: Starting Exit the Session...
14:52:42 mail systemd[1907]: Stopped target Default.
14:52:42 mail systemd[1]: Stopping Session 1 of user gavannon.
14:52:42 mail systemd[1907]: Stopped target Basic System.
14:52:42 mail systemd[1907]: Stopped target Sockets.
14:52:42 mail systemd[1]: Stopped target Graphical Interface.
14:52:42 mail systemd[1907]: Stopped target Paths.
14:52:42 mail systemd[1907]: Stopped target Timers.
14:52:42 mail systemd[1]: Removed slice system-ifup.slice.
14:52:42 mail systemd[1]: Removed slice system-systemd\x2dfsck.slice.
14:52:42 mail systemd[1]: Stopped target Timers.
14:52:42 mail systemd[1]: Stopping Accounts Service...
14:52:42 mail systemd[1]: Stopped target Multi-User System.
14:52:42 mail systemd[1]: Stopped Set Cloud Password.
14:52:42 mail systemd[1]: Stopping LSB: Record successful boot for GRUB...
14:52:42 mail systemd[1]: Stopping LSB: Starts amavisd-new mailfilter...
14:52:42 mail systemd[1]: Stopping OpenBSD Secure Shell server...
14:52:42 mail systemd[1]: Stopped target Login Prompts.
14:52:42 mail systemd[1]: Stopping Getty on tty1...
14:52:42 mail systemd[1]: Stopping LSB: XenServer Virtual Machine daemon providing host integration services...
14:52:42 mail rsyslogd: [origin software="rsyslogd" swVersion="8.12.0" x-pid="729" x-info="http://www.rsyslog.com"] exiting on signal 15.

(Something must be happening here, because the next line isn't for 30 minutes)

15:23:04 mail rsyslogd: [origin software="rsyslogd" swVersion="8.12.0" x-pid="770" x-info="http://www.rsyslog.com"] start
15:23:04 mail rsyslogd-2222: command 'KLogPermitNonKernelFacility' is currently not permitted - did you already set it via a RainerScript command (v6+ $
15:23:04 mail rsyslogd: rsyslogd's groupid changed to 109

这种情况发生在14:52的最后一件事是日志关闭（rsyslogd说exiting）。下一行是在启动时再次启动的日志，但是30分钟之后！