All virtual machines cannot be managed from PACI on one particular PCS node. The following symptoms are possible:
On attempting to stop a virtual machine it gets stuck in stopping state.
New virtual machines cannot be created — it gets stuck in "Creation in progress" state.
On attempting to open console session to a virtual machine it fails with error:
Timeout for console initialization has expired.
On hardware node
parallels-serverservice may become dead:
~# /etc/init.d/parallels-server status parallels-server dead but subsys locked
A lot of
stunnelerrors are logged in
F /disp:823115:833017/ Error: the stunnel has quitted unexectedly
The following version of
stunnelpackage was installed:
~# rpm -q stunnel stunnel-4.29-3.el6_6.1.x86_64
Due do incompatibility with the new version of
stunnel package provided by CloudLinux, Parallels Dispatcher Service may misbehave.
The issue was fixed permanently since Parallels Cloud Server 6.0 Update 9 (6.0.9-2806) in scope of the bug with internal ID PSBM-31110.
Install all the pending updates and reboot into the latest vzkernel.
In case installing updates was postponed, the
stunnel package should be downgraded.
Download and install
~# rpm -Uvh --oldpackage http://updates.cloudserver.parallels.com/cloudserver/6.0/Release/Packages/stunnel-4.29-3.el6_4.x86_64.rpm
stunnel package from yum repositories:
~# sed -i '/psbm6-updates.mirrorlist/ a\exclude=stunnel*' /etc/yum.repos.d/cloudlinux-parallels.repo
Restart Parallels Dispatcher service.
NOTE: restarting Dispatcher is required and it will restart all Virtual Environments (both Containers and Virtual Machines)
~# /etc/init.d/parallels-server restart
The workaround should be reverted before updating to Parallels Cloud Server 6.0 Update 9 (6.0.9-2806), if
stunnel was excluded globally from all repositories:
125156 PACI suspended or stopping virtual machines after updating to PCS 6 update 9