cancel
Showing results for 
Search instead for 
Did you mean: 

GCO heartbeat has status Exited Icmp unknown

tunisGharbi
Level 4
Partner Accredited

After moving server to secondary site, replication has restarted and working - good - except the Icmp status between gco  cluster: status  Exited Icmp unknown

 

Please Note that the distance between the 2 sites is 500km, the bandwidth is about 1.5mbs

and the following tunig parameters:

heartbeat Icmp (
  ...
    AYAInterval @vcs-remote = 0
    InitTimeout @vcs-remote = 0
    StartTimeout @vcs-remote = 0
    CleanTimeout @vcs-remote = 0
    StopTimeout @vcs-remote = 0
    AYATimeout @vcs-remote = 0
    AYARetryLimit @vcs-remote = 0
    Arguments @vcs-remote = { "x.x.x.169", "" }

please any fedback regarding   hahb -display
Icmp         State          UNKNOWN

Regards

 

2 REPLIES 2

mikebounds
Level 6
Partner Accredited

Why have all the variables in the heartbeat been set to zero - have you tried with the default values.  I would remove heartbeat and re-add.  If this works, then you can try tuning attributes by reducding their values if required, but they should not be zero.

Mike

starflyfly
Level 6
Employee Accredited Certified

refer :

 

https://iconnect-symwise.symantec.com/infocenter/index?page=content&id=TECH33539&locale=en_US

 

Problem

Icmp heartbeats show State "UNKNOWN" in hahb -display with Global Cluster Option (GCO)

Solution

The Icmp heartbeat may display a State of "UNKNOWN", even though the ClusterAddress for the remote cluster has been configured.

node# hahb -display
#HB          Attribute     Scope          Value
Icmp         AYAInterval   global         60
Icmp         AYARetryLimit global         3
Icmp         AYATimeout    global         300
Icmp         AgentState    global         ON
Icmp         Arguments     clus_remote    10.70.106.28
Icmp         CleanTimeout  global         300
Icmp         ClusterList   global         clus_remote
Icmp         InitTimeout   global         300
Icmp         LogDbg        global
Icmp         StartTimeout  global         300
Icmp         State         clus_remote    UNKNOWN
Icmp         StopTimeout   global         300

Here, we can ping the IP:

node# ping  10.70.106.28
10.70.106.28 is alive

The solution is to place the remote cluster's ClusterAddress IP and hostname in DNS or in the local /etc/hosts file:

node# more /etc/hosts
#
# Internet host table
#
127.0.0.1       localhost
10.70.106.12    node  loghost
10.70.106.28    clus_remote

This example includes only one failover cluster.  Your configuration may be different.

After putting this information in the hosts file, the Icmp heartbeat has the proper ALIVE state:

node# hahb -display
#HB          Attribute     Scope          Value
Icmp         AYAInterval   global         60
Icmp         AYARetryLimit global         3
Icmp         AYATimeout    global         300
Icmp         AgentState    global         ON
Icmp         Arguments     clus_remote    10.70.106.28
Icmp         CleanTimeout  global         300
Icmp         ClusterList   global         clus_remote
Icmp         InitTimeout   global         300
Icmp         LogDbg        global
Icmp         StartTimeout  global         300
Icmp         State         clus_remote    ALIVE
Icmp         StopTimeout   global         300