Skip to main content

7. Starting and Stopping VCS

Starting LLT and GAB

First, make sure that llt and gab are running on all nodes in the cluster.

# lltstat -n
LLT lltstat ERROR V-14-2-15000 open /dev/llt failed: No such file or directory

We can see llt is not running, so need to start it and check it again. This needs to be done on all nodes.

# service llt start
Starting LLT:
LLT: loading module...
Loaded 2.6.18-8.el5 on kernel 2.6.18-128.el5
WARNING: No modules found for 2.6.18-128.el5, using compatible modules for 2.6.18-8.el5.
LLT: configuring module...
# lltstat -n
LLT node information:
Node State Links
* 1 vmlinux3 OPEN 3

Once llt is running on all nodes we need to make sure that GAB is running.

# gabconfig -a
GAB gabconfig ERROR V-15-2-25022 unknown error

We can see gab is not running so need to start it

# service gab start
Starting GAB:
Loaded 2.6.18-8.el5 on kernel 2.6.18-128.el5
WARNING: No modules found for 2.6.18-128.el5, using compatible modules for 2.6.18-8.el5.

Check the status afte starting it. In the example below this is the first node and hence we do not have Port A membership yet.

# gabconfig -a
GAB Port Memberships
===============================================================

Once llt and gab are running on all node each node should report Port A membership as below.

# gabconfig -a
GAB Port Memberships
===============================================================
Port a gen c98301 membership 01

Starting VCS

Once LLT and GAB are runinng on all nodes we need to start VCS itself.  The following needs to be performed on all nodes.  The first node to be started will read its configuration from disk.  The next nodes to join will read the configuration from the existing cluster nodes via GAB.

Below we start had on the first node.

# hastatus -sum
VCS ERROR V-16-1-10600 Cannot connect to VCS engine
VCS WARNING V-16-1-11046 Local system not available

VCS (the had daemon) is not running so we need to start it.

# hastart

Wait a few seconds for the daemon to initialise then check the status. We should also see the had and hashadow daemons.

# hastatus -sum

-- SYSTEM STATE
-- System State Frozen

A vmlinux2 RUNNING 0
A vmlinux3 UNKNOWN 0
# ps -ef | grep had
root 7093 1 0 14:47 ? 00:00:00 /opt/VRTSvcs/bin/had
root 7096 1 0 14:47 ? 00:00:00 /opt/VRTSvcs/bin/hashadow

The above was run on the first node which is why vmlinux3 is in an unknown state.  Now run hastart on vmlinux3 to bring it into the cluster.

# hastatus -sum
VCS ERROR V-16-1-10600 Cannot connect to VCS engine
VCS WARNING V-16-1-11046 Local system not available

VCS is not running so start it.

# hastart

Wait a few moments

# hastatus -sum

-- SYSTEM STATE
-- System State Frozen

A vmlinux2 RUNNING 0
A vmlinux3 RUNNING 0

Finally, on all nodes in the cluster run gabconfig -a again to verify Port H membership.

# gabconfig -a
GAB Port Memberships
===============================================================
Port a gen c98301 membership 01
Port h gen c98304 membership 01

Stopping VCS

To stop VCS on all nodes and offline all running Service Groups:

# hastop -all 

To stop VCS (had) but leave the service groups running

# hastop -force

To stop VCS on the local machine and evacuate service groups

# hastop -local -evacuate

after running the above the othe node will show this nodes status as 'EXITED'

# hastatus -sum

-- SYSTEM STATE
-- System State Frozen

A vmlinux2 EXITED 0
A vmlinux3 RUNNING 0