Error while installing Ambari

I am using centos 5.* to install ambari not sure why it is looking for centos64 here.

17:37:42,956 ERROR AmbariServer:399 – Failed to run the Ambari Server
com.google.inject.ProvisionException: Guice provision errors:

1) Error injecting constructor, java.net.UnknownHostException: centos64: centos64
at org.apache.ambari.server.bootstrap.BootStrapImpl.(BootStrapImpl.java:53)
at org.apache.ambari.server.bootstrap.BootStrapImpl.class(BootStrapImpl.java:45)
while locating org.apache.ambari.server.bootstrap.BootStrapImpl

Topic

[RobertMember]

Hi Serge,
Definitely explore the time on the boxes. If they are not properly synched, issues can occur. In addition, does the output of hostname -f match of what you put in the host list for Ambari to register? For instance if hostname -f has the output of test.com, you should put test.com within the ambari hosts list section during the installation. In addition, you can also try running this api command in your browser to verify if the host actually registered into ambari:

NOTE: make sure you have logged in as admin successfully first into your browser prior to going to the below url:

http://:8080/api/v1/host

Place the above url within a browser and an output like below will display:

{
“href” : “http://192.168.56.101:8080/api/v1/hosts”,
“items” : [
{
"href" : "http://192.168.56.101:8080/api/v1/hosts/localhost.localdomain",
"Hosts" : {
"host_name" : "node1.example.com"
}
}
]
}

This command should let us know if it is getting to the point of registering.

Hope that helps.
Regards,
Robert

[Serge DeragonMember]

Hi Robert,
Sorry for the delay, been out&about…
Yes i did get the right files for the OS.

Hi Larry,
Seems this one node/one server thing is not going to be easy.
As sysadmin, been moving things around and making links to satisfy your install process.
But i’m at a lost here again.
I’ve managed to logon to the ambari host:8080 and bring up the installer.
After giving it what it needs(ssh keys and jdk home) i still get a failed at the Confirm Hosts sections.
The UNAVAILABLE stuff is there I.E ntpd running, not all others running, think there’s a variable i should exploit.
Below is most of the text. Much obliged for the time you are taking to educate me.
BTW, my associates want to know what it takes to be partners.
*******
STDERR
STDOUT
Verifying Python version compatibility…
Using python /usr/bin/python2.6
Checking for previously running Ambari Agent…
ERROR: ambari-agent already running
Check /var/run/ambari-agent/ambari-agent.pid for PID.
(‘hostname: ok e2i1.esound
ip: ok 192.168.1.104
cpu: ok Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz
Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz
Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz
Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz
memory: ok 7.39055 GB
disks: ok
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg_e2i1-lv_root 50G 5.7G 42G 13% /
tmpfs 3.7G 232K 3.7G 1% /dev/shm
/dev/sda1 485M 38M 422M 9% /boot
/dev/mapper/vg_e2i1-lv_home 53G 180M 51G 1% /home
os: ok CentOS release 6.4 (Final)
iptables: ok
Chain INPUT (policy ACCEPT 0 packets, 0 bytes)
pkts bytes target prot opt in out source destination
Chain FORWARD (policy ACCEPT 0 packets, 0 bytes)
pkts bytes target prot opt in out source destination
Chain OUTPUT (policy ACCEPT 0 packets, 0 bytes)
pkts bytes target prot opt in out source destination
selinux: ok SELINUX=disabled
SELINUXTYPE=targeted
yum: ok yum-3.2.29-40.el6.centos.noarch
rpm: ok rpm-4.8.0-32.el6.x86_64
openssl: ok openssl-1.0.0-27.el6.x86_64
curl: ok curl-7.19.7-35.el6.x86_64
wget: ok wget-1.12-1.8.el6.x86_64
net-snmp: ok net-snmp-5.5-44.el6.x86_64
net-snmp-utils: UNAVAILABLE
ntpd: UNAVAILABLE
ruby: UNAVAILABLE
puppet: UNAVAILABLE
nagios: UNAVAILABLE
ganglia: UNAVAILABLE
passenger: UNAVAILABLE
hadoop: UNAVAILABLE
yum_repos: ok
HDP-UTILS-1.1.0.15 Hortonworks Data Platform Utils Version – HDP-UTILS-1. 52
zypper_repos: UNAVAILABLE
‘, None)
(” u’responseId’: 23,
u’restartAgent’: False,
u’statusCommands’: []}
INFO 2013-04-02 15:15:51,097 Controller.py:116 – No commands from the server : []
INFO 2013-04-02 15:15:51,097 Controller.py:116 – No commands from the server : []
INFO 2013-04-02 15:15:51,097 Controller.py:180 – No commands sent from the Server.
INFO 2013-04-02 15:16:01,103 Heartbeat.py:68 – Heartbeat dump: {‘componentStatus’: [],
‘hostname’: ‘e2i1.esound’,
‘nodeStatus’: {’cause’: ‘NONE’, ‘status’: ‘HEALTHY’},
‘reports’: [],
‘responseId’: 23
Etc…t’ill registration with server fail

[Larry LiuModerator]

Hi, Serge

Thanks for letting me know.

Larry

[Serge DeragonMember]

Hi Larry,

Resolved.
With your insight, figured out that CentOS 6.4 needs some specifics to know what it’s hostname should be.
Probably the same for other Linux flavors.
I’m used to Solaris, HPUX and AIX, no need for so many things.

Took no chances and :
Put the hostname in /etc/hostname. Created it since CentOS did not at install, not sure if or why it uses it because no result with only that part.
Put the hostname, ip and gateway in /etc/sysconfig/network
Put the ip and short hostname in /etc/hosts

Since i installed the server with dhcp, made sure that the router/dhcp server reserved the same ip for the machine’s mac.
That way, everybody happy including DNS and probably DA(ldap) at reboot.
Ambari-server start and VOILÀ.

Thx again, continuing.

[Serge DeragonMember]

Larry Liu wrote:

Hi, Serge,

Are you testing a single node cluster? Here are a few things to check:

1. check ‘hostname’ and ‘hostname -f’ and make sure they are identical.
2. Since you use dhcp, please make sure you add ‘ip hostname’ to /etc/hosts

Please let me know how’s thing going.

Thanks
Larry

Hi Larry

Thx for your quick response.
Will implement.
Your site and expertise are lifesavers.

I’m testing with one node and will be building a setup for a minimum of 3 nodes for clients.
Most likely clients with private clouds (OpenStack Rackspace) or other for data and backups.

The plan is a key in, blind eyed, flexible bigdata and backups setup, using/promoting your tech and knowhow for my Canadian and European clients.

Most of my clients are French speaking.

Will keep you in the loop.

A+

[Larry LiuModerator]

Hi, Serge,

Are you testing a single node cluster? Here are a few things to check:

1. check ‘hostname’ and ‘hostname -f’ and make sure they are identical.
2. Since you use dhcp, please make sure you add ‘ip hostname’ to /etc/hosts

Please let me know how’s thing going.

Thanks
Larry

[Serge DeragonMember]

Hy guys,
I’m to be the guru for an international company that intends to user hadoop for bigdata.
I really would love to look like a hero rather than a zero LOL.

While doing my first install , i have the same problem with CentOs 6.4. with local dhcp setup on a brand new pc server with 8gb of ram.
After several reinstall successfull/without errors, i get ambari-server start Error in the /var/log/ambari-server/*.log is :
Tried deleting /var/lib/pgsl and reinstalling Ambari server.
ERROR AmbariServer:395 – Failed to run the Ambari Server com.google.inject.ProvisionException: Guice provision errors:
1) Error injection constructor, java.net.UnknownHostException: “myservername”: “myservername” at org.apache.ambari.server.bootstrap.BootStrapImpl.(BootstrapImpl.java:53) at…..

1 error
at com.google.inject.internal.InjectorImpl$4get(InjectorImpl.java:987)

And so on…

Could i prevail on you for some insight.

sderagon@gmail.com
Thanks

[RobertMember]

Hi Cloudack helpdesk,
We have different repos for centos 5 and centos 6 as listed here:

http://docs.hortonworks.com/HDPDocuments/HDP1/HDP-1.2.1/bk_using_Ambari_book/content/ambari-chap2-1.html

Can you confirm you downloaded the proper repo based on your operating system version?

Regards,
Robert

[Author]

Replies