Server fails on restart

[jmgorro]

Hi all

I have a Centos 7 box running 3.10.0-1160.76.1.el7.x86_64 that every time I try to shutdown or reboot it stucks on black screen (is a virtualized server). The only thing I can do is a power reset. Then it starts normally and runs fine with latest kernel.

When The server is in "black screen" the ping (from other networked computer) is still running but all other services are fully stopped. The server appears as running on hypervisor.

Having a look on /var/log/messages I can't see the power-off message. Last messages on log informs about stopping services.

I also waitied for 20 minutes to full shutdown but no luck (the server boots in few seconds so shutdown sounds to run on similar amount of time).

Also installed kernel-headers to solve Red Hat bug 1185918 (https://bugzilla.redhat.com/show_bug.cgi?id=1185918).

I have no idea where is the issue. Any help will be appreciated.

[jlehtone]

Could there be a race condition between some services so that they mutually wait for each other, without timeout?
Does log have "Stopping" without "Stopped" for some services?
Are there "Started" services that never log "Stopping"?

[jmgorro]

Hello jlehtone and thanks for your reply.

I've tried to reboot server again and, once restarted, checked one by one all services that weren't stopped. I found 3: Samba, Clamd and Postfix.

All other Stopping requests has their corresponding stopped log line.

I tried to stop them manually (no issues found) and then restart the server again. But same behavior. In this case only 4 Stopping requests have been launched (as registered on log), anyone has their stopped pair:
* LVM2 PV scan on device 8:2...
* Service for virtual machines hosted on VMware...
* Network Time Service...
* System Logging Service...

Searching all this servides in previuos log I can see all of them becomes stopped except last one. The "System Logging Service" does not appear on previous logging.

Still confused.

Thanks.