AnsweredAssumed Answered

Why CLDB service is not running on Control System WEB UI?

Question asked by zoglee on Sep 18, 2011
Latest reply on Sep 19, 2011 by yufeldman
hi, I'm very interested in MapR and I install a cluster(M3).

The deployment-plan is: master(1) + slaves(5):

> master: CLDB, jobtracker, fileserver,
> zookeeper, webserver, nfs
>
> slave1: fileserver, tasktracker,
> zookeeper
>
> slave2: fileserver, tasktracker,
> zookeeper
>
> slave3 to slave5: fileserver,
> tasktracker

It seems everything is fine, but I just find a strange phenomenon on the Control System WEB UI:

> Services
>
> **CLDB         0 of 1 instances running**
>
> FileServer 5 of 6 instances running
>
> JobTracker **0** of 1 instances
> running
>
> TaskTracker 5 of 5 instances running
>
> NFS Gateway **0** of 1 instances
> running
>
> HostStats **5** of 6 instances running
>
> **Webserver 0 of 1 instances running**

But, I can see all of the services on the CLDB-node with `jps' command:

> root@master:~# jps
>
> 22985 QuorumPeerMain
>
> 24170 CommandServer
>
> 20872 Jps
>
> 11998 Bootstrap
>
> 23477 CLDB
>
> 24085 JobTracker
>
> 23086 WardenMain

and the cluster seems do all the things correctly, such as hadoop shell command, mapreduce jobs, etc.

I just wanna know: is there anything matter or risk with my cluster?

(There are too many log files, and I couldn't tell which one is helpful as checking them all by myself, if you think anything else will be helpful, please ask me to show them out.)

Outcomes